We are living under a postdigital condition: a state in which digital media are no longer experienced as separate, novel, or even particularly visible—but instead fully integrated into the material and social fabric of everyday life.

Meanwhile the 2026 Iran War is unfolding through AI-generated images, videos, and memes — shaping how this new kind of AI war is experienced. Building on the last edition (Cognitive Warfare) here is a look at current developments with a postdigital propaganda lens.

Subscribe now

What the heck is Postdigital Propaganda™

Postdigital propaganda refers to the infrastructural organisation of political alignment within algorithmically governed platform ecologies. Yeah, I know. In plain terms: it means propaganda is no longer primarily about messages, lies, or even persuasion. It’s about the conditions under which things become visible, feel right, and get endlessly repeated.

The shift looks something like this: from message to environment, from actor to infrastructure and from persuasion to alignment. Alignment is when the environment itself starts to lean — as platforms consistently surface, reward, and circulate content pointing in the same direction, regardless of who produced it or why.

It’s called postdigital because we are no longer “online.” The Internet, Google, platforms. They are just there. Ambient. Like air. You only notice them, when they are gone. When platforms become the environment, propaganda doesn’t just move through them—it emerges from how they work.

So instead of digital propaganda = actors using tools. We get postdigital propaganda = alignment produced inside the system itself. Propaganda becomes a condition, not a campaign.

Four dynamics keep this running

You can see this everywhere right now: White House memes, Lego missiles, Ghibli soldiers, etc. It’s always the same four dynamics. These don’t run in sequence — they reinforce each other:

• Platform infrastructure: what you see is determined by a recommendation system — not the sender, not the argument.

• Platformized affect: what gets recommended is determined by reaction — outrage, delight, disgust. Feeling is the ranking signal.

• Participatory recursion: what circulates is continuously remade — by users, by networks, by state actors.

• Synthetic plausibility: what repeats starts to feel real — AI and aesthetics don’t need to convince, only to accumulate.

The Iranian Lego Use Case

Take the Lego propaganda videos making the rounds (How Lego Became a Go-To Meme of the Propaganda Wars, Wall Street Journal). On the surface, it’s absurd: missile strikes, dead soldiers, geopolitical escalation — rendered as cute plastic toys. But that’s exactly why it works. It runs on four dynamics — all at once.

Platform infrastructure: recognizable, frictionless formats — Lego, Minecraft, GTA — fit the feed and travel regardless of who made them. Which matters, because who made them is increasingly beside the point. The White House is running the same playbook, splicing real airstrike footage with Call of Duty overlays and Wii graphics. Iran didn’t infiltrate a neutral medium. The format was already there. What looks like foreign propaganda is structurally indistinguishable from official messaging — authorship collapsed into a shared visual vernacular no one owns.

Platformized affect: the childlike aesthetic doesn’t just lower your guard — it registers as an affective signal before you’ve evaluated anything. The system rewards resonance, not accuracy. Emotional charge determines visibility before origin or intent ever enter the picture. It doesn’t spread because it’s true. It spreads because it hits. It doesn’t arrive pre-disarmed. It arrives pre-amplified.

Participatory recursion: Iran, Russia, China, and the White House remix the same visual language — but so do millions of anonymous accounts that aren’t coordinating with anyone. That’s the point. The format reproduces across actors until the distinction between propaganda and posting collapses. The system processes all of it the same way, indifferent to intent. It stops belonging to anyone. It becomes the condition.

Synthetic plausibility: the mechanism isn’t deception — it’s accumulation. Repetition stabilizes a political world before any single claim is evaluated. The game aesthetic doesn’t just make violence digestible. It reframes what this war is. Not tragedy, not crisis — content. Something to watch, remix, scroll past. By the time you see it, the imaginary already holds. Not because you’ve been fooled, but because it already feels coherent.

That’s the point: the video doesn’t need to convince you. It needs to circulate, register, repeat. And once it does, it stops being a clip and becomes part of the environment in which the war is understood — not as competing narratives, but as a shared affective atmosphere that everyone feeds and no one controls.

What changes

Under these conditions, the success of propaganda is no longer determined by whether something is true, or even persuasive. Postdigital propaganda does not just shape what we see or feel—it also shapes what counts as evidence, and how reality itself can be confirmed or denied (Tech Policy Press) as long as it

• fits the feed

• hits the feeling

• gets picked up

• and keeps looping.

So what

We can keep arguing about what is true and what is fake. Keep chasing the next debunk, the next fact-check. But that’s increasingly beside the point.

If the problem is alignment, the task shifts: away from correcting individual claims — and toward understanding, and intervening in, the conditions that make certain directions stick.

Disinformation assumes a problem of content. We’re dealing with a problem of condition.

More
Cascades of A.I. Fakes (New York Times)
The fog of AI (The Atlantic)
Trump’s video game war: AI, memes and a simplistic narrative (Guardian)
War Becomes Spectacle in Trump’s Horrific Propaganda Promoting War in Iran (Truthout)

What else?
Three-quarters (73%) of Gen Z TikTok active users say the content today feels staged and performative, and over half (53%) feel it’s censored (The Harris Poll)

Finding fake followers, TikTok edition (Conspirador)
The material life of Italian Brainrot (Algofolk)
The Story Behind All Those Fruit and Veggie AI Slop Videos (NY Mag)
AI videos of sexualised black women removed from TikTok (BBC)
Generative Propaganda (Working Paper)

Thanks for reading. My name is Marcus. If you’re interested in tailored insights, workshops, consulting, or policy support, or just want to discuss stuff get in touch. Here is Linkedin. Here is Bluesky. Ciao

Subscribe now

The 2026 CogWars

Back in December 2025 NATO published a Chief Scientist’s Report on Cognitive Warfare (love the short form: CogWar). CogWar builds up, but is distinct, from older concepts like psychological operations or information warfare.

CogWar basically seeks to “exploit facets of cognition to disrupt, undermine, influence, or modify human decision-making by altering human behaviour and cognition through any means and technological advances.” It uses military and non-military tactics, targeting both military operators and civilian populations. Yeah, that is us.

It is worth mentioning that every party to this war — the US, Israel, Iran — and a couple of others appear to be running their own information operations simultaneously. The polluted, disordered information spaces we are in are currently extra-loaded.

All of us are right now being subjected to competing cognitive influence campaigns from multiple directions, making it genuinely difficult to form a clear picture of what is actually happening. Here are just two random examples of war related content on TikTok’s For You these days.

Under postdigital conditions, the efficacy of any strategic intervention is increasingly determined by its alignment with the algorithmic, affective, participatory and synthetic logics of the infrastructural environment rather than by the quality of its arguments or the accuracy of its claims. This perspective helps understand videos like this one, posted by the official White House TikTok account on March, 5, 2026.

While U.S. military investigators believe it is likely that U.S. forces were responsible for an apparent strike on an Iranian girls’ school that killed scores of children on Saturday (Reuters) the White House posted a video utilizing the Lizard Lizard Lizard Meme (Know Your Meme), one of the top trending memes on Google of 2025.

Just as with the Nihilistic Penguin (#150) there are several layers to understanding this meme. The meme comes from a post-credits teaser released at the end of the 2025 Disney film Elio to promote the 2026 Disney Pixar film Hoppers. Jokes on TikTok typically revolve around doing a repetitive and annoying task, put here in the context of launching rockets.

While the dominating war aesthetic used to be coined by computer games, we are witnessing another iteration of mimetic cutification. Compare the ribbon emoji (”often used to indicate something is pretty, cute, or special. Also used of various content concerning baby girls“) on the B2-bomber above. The Lizard has a Pixar aesthetic, compare that to the IDFs or Trump administrations usage of the Ghibli filter.

Obviously there are other potential layer to the Lizard. I just point you to this NBC article: Like QAnon’s Capitol rioters, the Nashville bomber’s lizard people theory is deadly serious Bonkers? Sure. Harmless? Definitely not. Trump has signaled affinity with QAnon followers in social media post, at rallies before (CBS).

Coming back to CogWars: It is not the deciding question wether the White House is signaling to QAnon here, but that they could and would. It isn’t just about messaging. It is how people process information. Videos like these do not serve a single target or outcome as CogWar can be designed simply to create confusion. Or for the LOLz, the vibes, the giggles, the shrugging, the outrage. Flooding the zone. Feeding the algorithms.

Now what? We can immerse ourselves in endless streams of horrendous videos or try to increase our resistance in order to being better equipped to withstand, recover, and continue functioning somewhat clearly despite the cognitive assaults.

Postdigital Propaganda

I have two new papers out (Open Access). Both analyse aspects of postdigital propaganda. I am in the midst of tying them and two more together under the umbrella term.

Bösch, M. (2026). TikTok edits, vibes, audio memes: Participatory propaganda in the 2025 German federal election campaign. Media and Communication, 14. https://doi.org/10.17645/mac.11755

Bösch, M., & Divon, T. (2026). Dancing in Uniform: The Role of Thirst Trap Propaganda in Military Image Wars on TikTok. Platforms & Society

What else

Pew Research Center: 8 facts about Americans and TikTok
Bellingcat: Viral Child Soldiers on TikTok: The ‘Disney Stars’ of Sudan’s Civil War
Reuters: TikTok charged for breaching EU rules with app’s addictive features
EUObserver: TikTok used automation in nearly 100% of violating-content moderation
Social Media & Society: The TikTok Caliphate: How Jihadist Supporters Exploit
Nature: The science influencers going viral on TikTok to fight misinformation

Thanks for reading. If you’re interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch. Here is Linkedin. Here is Bluesky. Ciao

Subscribe now

🐧 The Nihilist Penguin — a Meme Nazis Love to Appropriate

Fox News got it right: President Trump’s penguin post (that you have most likely seen) wasn’t ignorant; it was a deliberate nod to a viral right-wing meme (Fox News).

On January 23rd, 2026, the official White House Twitter/X and Instagram accounts shared a Nihilist Penguin meme that depicted an AI-generated photo of a penguin holding an American flag next to U.S. President Donald Trump with Greenland’s flag planted in the background near mountains with the caption, “Embrace the penguin.” The post referenced the Trump administration’s push to acquire Greenland that was prominent at the time of the meme’s circulation. Later on January 23rd, the official DHS Twitter account (DHSgov) also shared a version of the meme with the caption, “Americans have always known ‘why,’” receiving over 5.1 million views, 41,000 likes and 2,300 replies in three days (Know Your Meme).

(Source: DHS)

While parts of the internet (including me and a lot of commenters underneath the official Insta post) were still shaking their heads at yet another episode of presidential AI-slop nonsense – after all, penguins live at the South Pole – the internet’s content-exploitation machine was already in full mode. Here is an attempt to unwrap some layers of meaning.

📖 M — Meaning (Lore & Context)

☐ Do we understand the meme’s origin or background?
☐ Who is likely to understand it immediately — and who isn’t?
☐ Which parts of its meaning are coded, dogwhistled, or only visible to in-groups?

While the clip of the penguin from a Werner Herzog documentary 2007 was shared to TikTok as early as late 2024, it did not go viral and become a meme until mid-January 2026. On January 16th, 2026 an edit appeared on TikTok that included Werner Herzog’s narration and the organ cover of the song “L’Amour Toujours,” making it the first known example combining the two elements of the 2026 meme trend. The combination resonated well among users posting about e.g. “saving Europe” alongside quotes from German philosopher Friedrich Nietzsche (Know Your Meme).

“The online right saw the penguin as a powerful rebuke of secular modernity, marching toward a greater purpose” (Fox News). On January 23rd, 2026, the official White House Twitter / X and Instagram accounts shared an AI image referencing the penguin meme. On January 27, Lidl a trademark used by two German international discount supermarket chains used the footage (Herzog + Organ) for a video on TikTok. Other brands (Sixt, Deichmann) used it too.

Penguin: Emblem of masculinity

From the early days of internet culture, penguins have functioned as ideal meme animals: visually distinctive (upright, human-like posture), readable emotions (confusion, determination, isolation) and a strong contrast between cuteness and harsh environments. Compare the Socially Awkward Penguin (2009).

(Source)

The penguin first became an emblem of masculinity thanks to a previous viral video featuring a drag queen interrogating an elementary-school-age boy about men wearing makeup. The child asserted that boys cannot wear makeup. When the drag queen asked the boy, “Who said?” the boy pointed at a paper penguin on the wall and exclaimed, “The penguin over there!”. “The penguin over there” became a tongue-in-cheek internet defense of manhood and sex differences against the LGBT cult.” (Fox News)

Penguin + L’Amour Toujours

TikTokers paired the penguin footage with an organ remix (Youtube: Andreas Gärtner) of the right-wing anthem “L’Amour Toujours (I’ll Fly With You)” and overlaid images of Western heroes: Joan of Arc, Alexander the Great, Aragorn, Jesus Christ, King Baldwin IV and Luke Skywalker. Countless similar penguin edits have garnered millions of views online (Fox News).

From late 2023 onward, parts of the song’s melody were repurposed in Germany and elsewhere by far-right actors who attached explicitly racist chants such as “Deutschland den Deutschen – Ausländer raus” to the tune at parties and social media posts, including at a highly publicized incident on Sylt where such chanting and a Nazi salute were documented.

The appropriation sparked widespread condemnation: politicians denounced the behavior, copyright holders filed complaints, and events like the Oktoberfest banned the song to avoid further racist outbursts.

Geboers, M., & Bösch, M. (2025). Malicious earworms and useful memes, how the far-right surfs on TikTok audio trends (arXiv: 2506.20695). https://doi.org/10.48550/arXiv.2506.20695

The nihilist penguin meme variations using the L’Amour Toujours sound carry accumulated political meaning. Appropriated by far-right actors, the sound operates as a dog whistle shared by the Trump administration in the context of “publishing a stream of nazi propaganda” (The Atantic).

🎭 A — Affect (Vibe)
☐ What emotion does the meme trigger first?
☐ Is it earnest, ironic, cynical, or ambiguous?

This meme isn’t funny – it’s heavy. Especially due to German director Werner Herzog and his thick and uinque German accent that have helped him turn into a meme icon, used to signal seriousness, depth, and existential weight opposed to an internet culture otherwise dominated by irony, speed, and surface-level meaning. Meanwhile
the organ version of L’Amour Toujours replaces club nostalgia with ceremony, authority, and historical weight, adding legitimacy. The result is an affect that is openly earnest for insiders, affirming endurance and civilizational purpose, yet safely ambiguous to outsiders. It can be dismissed as AI absurdism – but it doesn’t feel that way to the people it’s meant for.

To get an idea, please read what Fox News has to say here: “DHS is right. America was built by penguins — and by that I mean rebels, pilgrims, frontier men and women, conquistadors, and cowboys. We are a nation founded by risk-takers who left the colony for the mountains. We are descended from men who suffered and died to carve civilization out of wilderness. It’s our inheritance. But beneath the idea of rugged individualism lies a deeper religious current. Some users have interpreted the mountains to symbolize Jesus Christ Himself.”

Given this affective and ideological context, the Lidl and German Federal Foreign Office adaptations risk reproducing a vibe whose political charge is already well established.

📱 T — Type (Format & Platform)

☐ What format is this (image, video, sound, gesture, multimodal)?
☐ Is it platform-native or recycled elsewhere?
☐ What role is AI playing here?

The Nihilist Penguin meme is a multimodal, platform-native TikTok format, built around a short video clip paired with music and minimal text overlays that function as a reusable affective template. While it emerges natively on TikTok, it travels easily across platforms, mutating into still images, screenshots, and political remixes on X, Instagram, and news media.

AI plays a secondary but enabling role, mainly through derivative image generation and rapid remixing, accelerating the meme’s circulation and allowing institutions and political actors to appropriate its recognizable visual language without engaging its original context.

⏳ T — Temporality (Lifecycle & Speed)

☐ Is this meme emerging, peaking, declining, or being reused?
☐ Are we early, on time, or late to the conversation?
☐ Has the meaning changed since it first appeared

The nihilist penguin meme has clearly moved past its organic TikTok peak and entered a phase of institutional reuse. What began in mid-January 2026 as a platform-native, far-right-coded remix format was rapidly consolidated through high-visibility amplification by the White House and DHS on January 23, marking a decisive shift from grassroots virality to top-down symbolic deployment. We are now in the phase of brand and institutional adaptations and variations signaling stabilization or the past-peak phase. We are late to the meme’s participatory moment but early in diagnosing its institutional afterlife. The window for organic remix culture has largely closed; what currently remains is a second life discussing the phenomenon.

⚖️ E — Ethics

☐ Is there any sign of manipulation, propaganda, or agenda-setting around it?
☐ Does this meme involve vulnerable people, trauma, or harassment?
☐ Could amplifying it cause harm, humiliation, or misinterpretation?

There is clear evidence of agenda-setting, even if not every instance constitutes deliberate extremist signaling. The White House and DHS posts mark a transition from subcultural meme use to state-level symbolic deployment, drawing on a meme already stabilized within far-right remix ecologies. The meme is structurally susceptible to dog-whistle readings because it combines elements that have accrued meaning through far-right appropriation. In this sense, the meme operates as ambient propaganda: not instructive or directive, but affectively aligning and normalizing a worldview through repetition and institutional endorsement.

The meme indirectly references symbolic violence through its association with exclusionary nationalist narratives and its musical lineage tied to racist chanting (“Deutschland den Deutschen – Ausländer raus”). The risk lies less in immediate harm than in normalizing semiotic materials that have been used to legitimize hostility. The danger lies precisely in its ambiguity, and its movement from fringe remix culture into institutional and commercial reuse where affect remains.

📈 R — Relevance

☐ Why does this meme matter now?
☐ Where is this meme circulating (which communities, regions, languages)?
☐ Who is amplifying it — individuals, influencers, media, coordinated actors?

This meme matters now because it has jumped from far-right remix culture into official state communication. It circulates across TikTok, X, and Instagram, slipping easily between languages, communities, and registers. What began in niche right-wing meme ecologies is now reused by brands, media outlets, and government actors, often stripped of context but not of meaning. Amplification no longer comes only from online subcultures, but from institutions whose endorsement normalizes the meme far beyond its original audience.

Hijacking Meaning in Dense Meme Ecologies

(Source)

Whether the German Federal Foreign Office successfully meaning-hijacks the meme is debatable. What is clearer is that its video – in parts – relies on the semiotic grammar and affective charge of the nihilist penguin, a meme already shaped by far-right remix cultures. This makes the message structurally vulnerable to misreading regardless of its stated intent. Especially when the dissimination of ideas relies on affect and vibe rather than rational discourse. Not sure if Macrons sunglasses, an animated Euro coin and a winking comic penguin are enough. The Edit for sure needs to be further dissected and analysed. Happy to hear your thoughts!


A Quote

“That’s AI” (It Isn’t) (The Present Age)

What else

US Power Play Over TikTok Did Nothing to Protect Americans (Tech Policy Press)
What the US TikTok takeover is revealing about censorship (The Guardian)
Gaza-based journalist Bisan Owda regains TikTok account after outcry (Al Jazeera)
Social media as state media (Popular Information)
The New Shadowbanning Panic (The Atlantic)
Khaby Lame and 𝘙𝘪𝘤𝘩 𝘚𝘱𝘢𝘳𝘬𝘭𝘦 𝘏𝘰𝘭𝘥𝘪𝘯𝘨 𝘓𝘵𝘥. (Linkedin)
TikTok users are deleting the app, with removals up 150% (CNBC)
TikTok blocks Epstein mentions and anti-Trump videos, users claim (Independent)
The Slopagandist (The Verge)

Thanks for reading. If you’re interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch. Here is Linkedin. Here is Bluesky. Ciao

Subscribe now

🇺🇸 The power to control a platform

“The TikTok deal encapsulates every abusive practice of the Trump administration—secrecy, unprecedented claims of executive power, and noncompliance with the law” (The Center for American Progress). TikTok is used by 170m Americans, including some 40% of 45- to 54-year-olds (The Economist).

TikTok has sealed a deal for new US joint venture to avoid American ban (Reuters). The newly formed “TikTok USDS Joint Venture LLC” will oversee the app’s operations, including data protection, algorithm security, content moderation, and software assurance (Techcrunch). “The Joint Venture will retrain, test, and update the content recommendation algorithm on U.S. user data. The content recommendation algorithm will be secured in Oracle’s U.S. cloud environment.” (Social Media Today).

“Now Collecting Even More Data”

Who owns TikTok in the U.S. now (New York Times). Learn more on Oracle’s Larry Ellison the “shadow president” (Truthout), the “Oracle-Palantir technology stack” and its role in Israeli military operations (Byline Times) and his broader project to reshape the American media (972 Magazine).

While it is still to early to assess the results of how the retraining of the algorithm will work out, the new privacy policy in the U.S (Last updated: January 22, 2026) shows the way. TikTok Is Now Collecting Even More Data About Its Users (Wired). It is worthwhile to compare the old to the new version.

Old: “Current versions of the app do not collect precise or approximate GPS information from U.S. users.”
New: “If you choose to enable location services … we collect approximate or precise location information from your device.”

Even if the changes seem small (like reducing a simple word like “only”). Although both versions of TikTok’s US privacy policy (prior and after January 22) list the same sensitive data — including religion, health, sexual orientation, gender identity and immigration status — the updated policy quietly changes who holds power over that information. Earlier language said TikTok would only process such data to operate the platform or under narrow legal exemptions. The new policy drops this limit and instead allows processing for broadly defined “permitted purposes” under California law.

This shift matters because it replaces clear, self-imposed constraints with wide legal flexibility, moving decisions about how deeply sensitive identity data is used away from users and regulators and into the platform’s own hands. In effect, the policy normalises the use of highly personal data as long as it is legally allowed, changing the governance logic from protecting users by default to granting TikTok greater discretion over what it can do with that data.

“The reality of domestic propaganda”

Governments are treating data and platforms as extensions of national influence, and market rules bend when politics and technology collide, Marina Yue Zhang and Wanning Sun write here. “My worry all along is that we may have traded fears of foreign propaganda for the reality of domestic propaganda”, says Georgetown law professor Anupam Chander (New York Times). A win-win for the Trump administration: A platform that collects vast amounts of most sensitive personal data in the hands of a friendly joint venture and yet another platform for propaganda and censorship.

China was just a proxy (Mashable). The whole US TikTok debate was always about infrastructural power. The power to control a platform that shapes public attention, culture, and political visibility.

🐸 Checklist: Evaluating Memes

I’ve created a small newsroom checklist for evaluating memes. It’s called M.A.T.T.E.R. — and it’s designed for those moments when a meme is everywhere and you’re unsure whether to cover it, contextualise it, or leave it alone.

📖 M — Meaning (Lore & Context)
🎭 A — Affect (Vibe)
📱 T — Type (Format & Platform)
⏳ T — Temporality (Lifecycle & Speed)
⚖️ E — Ethics
📈 R — Relevance

M.A.T.T.E.R. helps journalists make sense of memes without overreacting, under-explaining, or amplifying them by accident. It offers a fast, structured way to assess what a meme does, why it works, and whether it actually matters.

Use M.A.T.T.E.R. when a meme feels political, emotional, or risky — to slow down, separate feeling from meaning, assess impact and ethical risk, and make a conscious editorial decision.

I’ve stress-tested the framework on a mix of older and current memes, but I’d really love for others to try it out as well. Feedback, critiques, and edge cases are very welcome: m@marcus-boesch.de

📄 You can find a PDF version here, with additional guiding questions for newsroom use.

🍦 What else

The Trump Administration Is Publishing a Stream of Nazi Propaganda (The Atlantic)
The Great Entertainment (Kyla Scanlon)
TikTok’s ‘2026 Is The New 2016’ Trend, Explained (Forbes)
How China is battling the US online over Venezuela (DFRLab)
What Are “Fan Edits,” and Why Is Gen Z Obsessed With Them? (Inside Hook)
Can TikTok spark a new era of protests? (NPR)

Thanks for reading. If you’re interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch. Here is Linkedin. Here is Bluesky. Ciao

🪇 Macarena Militainment
🇻🇪 Maduro Memes
📷 The end of evidence

Subscribe now

🪇 Macarena Militainment

How do i explain to the normal person that the U.S. military is slowing down the song Macarena and using it for propaganda videos, asks TikTok user @dabootywizard. Actually the U.S military to my knowing is not running official accounts on TikTok different to e.g. Instagram (United States Army, US Navy). Yet there is a myriad of various private and inauthentic accounts (@u.s.airforce_offical with a typo) sharing U.S. military footage and confusingly TikTok offers AI generated overview pages for both Us Navy Official Account and the Us Military that indicate to present “authentic U.S. military experiences on TikTok”.

The most popular video of an account with the user handle @highaspectdesign posted a Macarena Edit on December 30, acquiring 5.3M views so far. From the comments: US navy soundtrack, If you hear this song in a foreign country it’s already over 😭.

Why Macarena

“Macarena” is a song by Spanish pop duo Los del Río, originally recorded in 1993 (Wikipedia) becoming one of the longest runs atop the Hot 100 chart in history, a harmless song overflowing with nauseating Latino-style beats, an easy choreography and repetitive lyrics about a girl named Macarena who cheats on her boyfriend, Vitorino, while he’s away serving in the military. Military? Wait, that is not the only potential reference here if you look closely.

Los del Rio were invited to tour South America in 1992 and, while visiting Venezuela, they were invited to a private party held by the Venezuelan impresario Gustavo Cisneros where a local flamenco teacher, Diana Patricia Cubillán Herrera, performed a dance for the guests, and Los del Río were pleasantly surprised by Cubillán’s dance skills.

Why Slowed Down

The slowed down reverb version has been used in military edits on Youtube for over a year. Slowing a track and adding heavy reverb makes it feel more nostalgic, dreamy, or melancholic, which matches a lot of TikTok edits.The spacey atmosphere created by reverb and lower tempo can make lyrics or melodies feel deeper or more dramatic and it fits the epic style of majestic warships and high-tech battle gear in an attempt to beautify war and the military (John Keane).

Why does it matter

The choice of sound makes sense in order to resonate in an affective vibed based surrounding where messages and moods perfectly blend in and become invisible propaganda. It is not just a harmless song, it is intended to spark a vibe utilized in the context of soft propaganda and militainment (Roger Stahl). It normalizes breaking international rules and bringing back the era of raw power.

📌 Unprovoked lethal strikes by the United States against vessels at sea may amount to international crimes (UN) –Dale a tu cuerpo alegría, Macarena 🎶

📌 We just witnessed power kidnapping the law (Professor of law at Howard University School of Law) – Que tu cuerpo es pa’ darle alegría y cosa buena 🎶

📌 President Donald Trump has called for US defence spending to be increased to $1.5tn (£1.1tn) in 2027 for what he called “these very troubled and dangerous times” (BBC) – Dale a tu cuerpo alegría, Macarena 🎶

📌 Trump really wants Greenland (FT) – Hey Macarena, ay! 🎶

🇻🇪 Maduro Memes

On Saturday, January 3 President Trump posted a photograph on Truth Social. “Nicolas Maduro on board the USS Iwo Jima,” Mr. Trump captioned the photo, minutes before he addressed the nation from Mar-a-Lago, his Florida resort, and announced that America was “going to run” Venezuela (New York Times).

The image is striking, strategically well placed and it tells us a lot about the U.S. administrations nimble communication strategy that is fast, memetic, and perfectly platformised. Taylor Lorenz has more on ICE’s social media strategy here.

Debbie Millman has written an excellent piece for Time about the New Visual Language of Power. Thankfully she contrasts the images to the ones of Saddam Hussein in 2003 that i instantly had to think of too.

Those photographs were engineered to communicate rupture and subjugation, writes Millman: disheveled hair, a visibly diminished body, the unmistakable signs of force. The message was explicit. Power had been stripped away, and the image insisted upon that assessment.

The photographs of Maduro operate differently, argues Millman. The Nike logo carries a distinct and widely shared set of meanings, often associated with athletes, celebrities, and leisure and the slogan “Just Do it” i may add. It functions not as a marker of authority, but as a mechanism of alignment... it collapses the distance that authoritarian imagery depends upon. Power ... is recontextualized.

The response to the photographs of Maduro has not inspired outrage or catharsis; instead, it has generated commentary, irony, and consumption, writes Millman. And it has sparked the usual cycle of memetic frenzy on TikTok and others. All those AI variations of Baby Maduro, DJ Maduro, Prisoner Maduro et al. all fuel the U.S. administrations goal: Normalise breaking international law for the LOLZ. While wrapping a foreign head of state into a comfy all-Amercian-tracksuit. Because we can. Raw power, wrapped in soft fleece / propaganda.

📷 The end of evidence

On January 7, 2026, Renée Nicole Macklin Good, a 37-year-old female US citizen, was fatally shot by a United States Immigration and Customs Enforcement (ICE) agent in Minneapolis, Minnesota. Despite claims made by the U.S. president, the U.S. Vice president, and the DHS Secretary that Good attacked ICE and attempted to run them over and ram them with her vehicle Bellingcat, The New York Times Visual Investigation Team, & Washington Post’s Visual Forensic team have all published analysis showing the ICE shooter wasn’t in the path of Renee Nicole Good’s vehicle when he shot her.

What is particularly striking in this case is the obvious and continuous lying by highest ranking offices of the U.S. government while the visual evidence is apparent. Vance: “She was trying to ram this guy with her car, he shot back, he defended himself... And everybody that is repeating the lie that this is some innocent woman who was out for a drive in Minneapolis, you should be ashamed of yourselves”. Tim Walz, governor of Minnesota and Democratic nominee for vice president in the 2024 U.S. presidential election, cited George Orwell (1984) in a public statement: “The Party told you to reject the evidence of your eyes and ears. It was their final, most essential command.”

Just hours after the deadly ICE shooting social media users began using AI to fill the gaps in the breaking news coverage to their own liking, writes Newsguard and points to various examples of false information generated with AI including a calm interaction between ICE agent and the victim. This further pollutes the information space and besides the information disorder it is one of the latest examples where “arguing over facts” has turned into “competing over vibes” as part of a vibocracy as explained in #147.

What else?

Together with an all-stars team of University Münster and HAW Hamburg i have worked on a paper “Synthetic disinformation detection among German information elites – Strategies in politics, administration, journalism, and business” that just got published in SCM - Studies in Communication and Media (Full Paper).

It argues that information elites from politics, administration, and business rely predominantly on internal strategies (i.e., something like “common sense”) to detect synthetic propaganda and AI-generated disinformation. Many cannot identify clear verification procedures—especially for fake and generated sound and audio snippets. Only the journalists interviewed were able to name external verification strategies. This again, relies on “arguing over facts”, right?!

Thanks for reading. Have a good day. Ciao.

you are reading Understanding Tiktok. My name is Marcus. 🚨 TikTok has finally signed a deal 🚨 to sell its U.S. business to three American investors — Oracle, Silver Lake and MGX — ensuring the popular social video platform can continue operating in the United States. Yeah, but how? The algorithms retrained on U.S. user data, the entire farce driven by geopolitical factors (#146) with more unhingend memes, slop, and dark signal circulation beyond surface level noise. Welcome to the vibocracy – where legitimacy is enacted rather than negotiated in volatile, performative arenas.

😵‍💫 From Vibe to Vib(e)-O-Cracy
🌊 Flood til you can’t flood no more
🎒 TikTok projects i worked on in 2025

Subscribe now

😵‍💫 From Vibe to Vib(e)-O-Cracy

Ever since Kyle Chayka’s TikTok and the Vibes Revival article from April 2021 i have used the term to describe, explain and understand TikTok. The great appeal always posed the greatest challenge as vibes serve as a placeholder for abstract qualities that you can’t pin down, representing a type of understanding that exists before words can even name the experience.

The term and its usage have evolved including a vibe shift (2022), a vibecession (2022), then corporate mainstreaming via vibe check (2023) and a “no thoughts, just vibes”-vibe, finally turning into vibeocracy (2024).

With vibeocracy P.E. Moskowitz in the summer of 2024 coined a political and social system where vibes replace traditional ideologies, policy, or objective facts with
the primacy of feeling over fact, aesthetic governance, and the end of shared truth. One year later Jacqueline Fendt (Emeritus Professor of Strategy and Entrepreneurship at ESCP Business School) got rid of the “e” and turned vibeocracy into vibocracy: a more formal sociological framework.

While P.E. Moskowitz describe the feeling of living in a vibe-driven world, Fendt describes the new system of government that has emerged because of it. You see Donald Trump hovering over the text as Fendt explains how societies move from “arguing over facts” to “competing over vibes”, where corruption is treated as a genre of performance, and where an audience accepts things as true, regardless of evidence if they feel right.

All of this happens in a “neo-oral” environment via audio-visual fragments (TikToks, memes, sound snippets) that vibe. A symptom and a product of contemporary media infrastructures where epistemic authority shifts.

Deep into the thoughtless hypnotism of TikTok one afternoon, I came upon an anonymous urban scene from inside a residential tower in Manchester, England. On an upper floor, beside huge windows shrouded in fog, a man was floating on his back in a gleaming pool, to the soundtrack of the plaintive Frank Ocean song “White Ferrari.” The ten-second video was hashtagged #vibin. Kyle Chayka in 2021.

White Ferrari was released in August 2016, three months before the 2016 US Presidental Election. But now it is 9 years on. 2025. The year we lived alt-2016 and debated everything.

🌊 Flood til you can’t flood no more

The British Economist and “America’s Most Trusted Dictionary” agree: Slop is the word of the year. The original sense of the word, in the 1700s, was “soft mud” now it is defined as “digital content of low quality that is produced usually in quantity by means of artificial intelligence. {#146} In 2024 we had Shrimp Jesus. In December 2025 we have:

• ‪TikTok spam accounts posting strangely similar AI-generated videos of nonexistent Black people (Conspirador Norteño)

• Ultrarealistic AI videos attempt to portray Ukrainian soldiers in peril (NBC)

• Thousands of AI videos featuring sexualised minors shared on TikTok (Euronews)

• Deepfakes of UK Prime Minister Flood TikTok (Newsguard)

• A16z-Backed Phone Farm Flooding TikTok With AI Influencers (404 Media)

• Agentic Al Accounts Flood TikTok With Harmful Content (AI Forensics)

• AI Slop for Babies (Bloomberg)

Now what?! China is cracking down on AI slop. The rules read like a shopping list of badly needed regulations in the US, writes Futurism.com. And while i embrace regulation in the cases mentioned above i’d politely point to Article 19 of the Universal Declaration of Human Rights1 and hesitate fully participating in another moral panic. By the way: Teenage Mutant Ninja Turtles take a stance against slop (Polygon) and so can you. Just turn the noise down (qz) or maybe a walk in the park (Beach House) when it’s cold outside (Erasure).

🎒 TikTok projects i worked on in 2025

#AI – Together with the cool team of AI Forensics i helped write a report on Agentic AI Accounts: Stanusch, N., Degeling, M., Buse Çetin, R., Bösch, M., Romano, S. (2025). Prompt, Upload, Repeat: How Agentic Al Accounts Flood TikTok With Harmful Content. AI Forensics https://aiforensics.org/work/agentic-ai-accounts

#Livestreaming – I joined the Digital Methods Summer School at University of Amsterdam to research TikTok livestreams and extremism: Van Boheemen, P., Bösch, M., Costanzo, G., Divon, T., Frühwirth, L., Hammelburg, E., & Wang, X. (2025). Live-Streaming: Mapping Networks of Influence and (Dis) information Flow. Digital Methods https://www.digitalmethods.net/Dmi/SummerSchool2025LiveStreaming

#Audio Memes – Together with Marloes i wrote a book chapter on malicious earworms, building up on previous research on audio memes: Geboers, M., & Bösch, M. (2025). Malicious earworms and useful memes, how the far-right surfs on TikTok audio trends (arXiv: 2506.20695). https://doi.org/10.48550/arXiv.2506.20695

#PoliticalCampaigning – For Friedrich Ebert Foundation i produced a research report about the German Federal Election 2025 on TikTok together with Jolan Geusen (German): Bösch, M., Geusen, J. (2025). Swipe, like, vote : Analyse des Bundestagswahlkampfs 2025 auf TikTok. Friedrich-Ebert-Stiftung e.V., Mai 2025. ISBN 978-3-98628-457-2

⚗️ In the pipeline: papers on thirst trap propaganda, synthetic propaganda and a closer look at audio memes, TikTok edits and vibes. Still out in the open: A meme curriculum for journalists and vague stuff including antimemes, vibocracy and the postdigital. Happy to discuss joint efforts and research proposals.

What else?

The Data Is In: The Washington Post Can’t Replace Its “TikTok Guy”. Legacy media can’t keep treating breakout creators as temporary assets or social experiments, they’re the connective tissue to new audiences. (Matt Karolian)

Politico has picked up on paneuropean propaganda, covered in #140 and #141

How TikTok works for digital diplomacy during conflict/war times: the cases of Russia-Ukraine and Israel-Gaza conflicts (Journal of Information Technology & Politics)

What makes a good social video? (Link in Bio)

Thanks for reading. If you’re interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch. Here is Linkedin. Here is Bluesky. Ciao

Subscribe now

🎰 Agentic AI Accounts

The term slop, a pejorative term for low-quality, high-volume generative AI content lacking effort or meaning, has already had an impressive career ever since 2022. We are now living a slop life (New York Times) coined by slopagada (Klincewicz, Alfano, & Fard) slopification (Tech Policy Press) and the slopverse (Ian Bogost), with browser tools like slop evader1 (404media) and a proposed tax on slop (Heavy Machinery). All this is happening in the midst of “the decline of the visual world of the modern age” as photographic artist Bernd Arnold argues.

Will we drown in nazislop (Garbage Day) and polarising engagement spam (404media) while the once open web “is increasingly a dead internet, inhabited by bots or — much worse — humans who have been conditioned into behaving like bots” (Kevin Munger)?

A new report by AI Forensics – that i had the pleasure to help with – sheds light on
so called Agentic AI Accounts2 (AAAs): How Agentic AI Accounts Flood TikTok With Harmful Content (PDF)

In our report, we focus on the malicious use of AAAs to maximise interaction and profit on human social media platforms by spreading engaging content, including plausible news of non-existent events, xenophobic and racist stereotypes, and sexualized (sometimes underage) female bodies. Despite posting harmful content AAAs often escape moderation for months and contribute to the fact that 55% of AI-generated content on platforms remains unlabeled, further emphasizing systemic weaknesses in platform governance.

Der Spiegel, leading German weekly news magazine and news website has covered the report here (german).

🦠 Dissecting a viral

Simon Pycha watches, collects and monthly ranks 136 German journalistic TikTok accounts for DWDL (November list). Besides he works as Social Media Analyst for 1Live, a radio station operated by the Westdeutscher Rundfunk public broadcasting corporation targeted at a younger audience (14-39) in North Rhine-Westphalia – the most populous state in Germany (18 million inhabitants). With 37.4M Likes 1Live (TikTok) is by far the most successful German “youth” “radio” channel on TikTok.

In a recent post on Linkedin Simon discloses some strategic decisions that led to a well-performing video (currently 2.1 M views), filmed during the German TikTok Awards 2025. As you can see it is not necessarily rocket science:

• Make the topic insider-focused to add real value and foster community engagement: What is your favourite niche account?

• Lead with your biggest name to hook viewers early, even if they don’t appear later (Nadine Breaty).

• Keep the intro and interviewer invisible — focus fully on the content.

• Structure clips strategically: strongest take first, another strong one last.

• Encourage interaction through audience guessing and call-to-action by a star for maximum impact.

• And: good editing

Try to spot these underlying strategic approaches when watching in order to better plan and execute your next video. If you must, that is. Maybe you prefer getting lost in the dark forest.

🔬 Researching TikTok

I was invited to present works and findings on Researching TikTok together with great colleagues like Julian Hohner (LMU München) and Lion Wedel (Weizenbaum Institut) at The Leibniz Institute for Media Research, Hans-Bredow-Institut (HBI).

Since the first ethnographic studies on TikTok so much has happened concerning methodologies. Yet naturally, there is no sense in trying to subsume all approaches to the study of apps under one comprehensive method (Duguay & Gold-Apel, 2023).

In order to unwrap never-ending streams of similar yet different content (Hagen, de Zeeuw, Venturini, 2025) and make sense of ephemeral phenomena i lean into the app’s hyper-individualization and personalization (Duguay & Gold-Apel, 2023) to study the complexities of such ‘sonic social media’. I want to point to a couple of papers that inform my way of thinking these days.

Make sure to check out: Digital Rhythmanalysis, Studying memetic and affective rhythms on the post-viral Web. I have included their brilliant Italian Brainrot Video Wall. You can do similiar things with the tool 4CAT.

📕 Bainotti, L., & Peeters, S. (2025). Enriching the metapicture: How to analyse audio-visual content on TikTok and other platforms. New Media & Society, 14614448251387109.

📕 Duguay, S., & Gold-Apel, H. (2023). Stumbling blocks and alternative paths: Reconsidering the walkthrough method for analyzing apps. Social Media+ Society, 9(1), 20563051231158822.

📕 Hagen, S., de Zeeuw, D., & Venturini, T. (2025). Digital rhythmanalysis: Studying memetic and affective rhythms on the post-viral Web. Platforms & Society, 2, 29768624251394967.

📕 Pilipets, E., & Chao, J. (2025). Noise in sonic social media: Memetic soundscapes of Deep TikTok. new media & society, 14614448251358752.

🗺️ What else?

Is the TikTok Sale Motivated by Geopolitical Factors? Yes, says International Banker: The TikTok sale is driven largely by geopolitical factors, particularly US concerns over controlling TikTok’s powerful algorithm and limiting Chinese influence amid US-China tensions. It highlights how key American investors involved in the deal have ties to Israel and framing the transaction as a strategic move to influence public opinion in the US, especially to restore Israel’s image after the Gaza conflict, rather than simply financial gain (International Banker) These echoes earlier findings, yet now you get them from a credible financial source.

Thanks for reading. If you’re interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch — I’d be happy to collaborate. Here is Linkedin. Here is Bluesky. Speak soon. Ciao. 🌲

Subscribe now

🛎️ The Great Meme Reset1

I have co-organised an interdisciplinary Meme symposium at LMU Munich together with my great colleagues Simon Lübke and Julian Hohner where i discussed the status quo of memes in late 2025. A special shout out to the ones who were there and are reading this. The entire day was a bliss.

The debate about the Great Meme Reset (Know Your Meme, Forbes, Wired, 8ball) reveals that “meme culture has reached its nostalgia era”. This basically means that memes – surprisingly – transform and have adapted to platform affordances. The image macro of the 2010s has been replaced by multilayered multimodal groups of items that are dense, ambigue, fluid and even harder to pin down, to research and to explain while they keep slipping through the fingers and keep changing the very meaning you were trying to grasp. Just think of the Pepe meme being used by nazis, the LGBTQI+ community, protesters in Hong Kong, lately re-appearing as Antifascist frog(s).

As top down slop memes reach “new levels of dehumanization” i call for a research agenda that is able to better unwrap rapid fire meme shitposting and military industrial complex memes by e.g focusing on the postdigital manifesting in the usage of gestures and sounds. Sorry, if this sounds rather cryptic. I am in the midst of elaborating on that in a more approachable way aka article. And i have a fancy slide set for a 30 minutes keynote. Book me.

🕵 50,000,000 AI videos can’t be wrong2

In the past 5 years since starting this newsletter i have most likely watched around 1 million videos on the platform. According to the Washington Post who collected a trove of more than 1,100 TikTok users’ viewing histories to dissect how people consume content on the social media platform.

Meanwhile TikTok has said that there are now 1.3bn videos on its platform labelled as AI-generated. As you might have seen by yourself most AI videos are not labeled. And most likely never will? Generative AI tools are extensively used to produce engagement spam, spread false information, and hateful and harmful content while the production and dissemination is more and more automated. Example: None of these people exist, but that hasn’t stopped them from posting profusely on TikTok. Another example: New AI video tools are fuelling violent racism on TikTok.

I recommend the study AI Generated Algorithmic Virality by the cool people of AI Forensics to learn more on that.

In Germany channels on TikTok are meanwhile inventing countless false government regulations and spreading them via generative AI videos. I discussed the phenomenon with the German Press Agency and public broadcaster Bayerischer Rundfunk. You can find articles here (German-Austrian Digital Media Observatory) and here (BR24). German though.

TikTok let’s you choose now how much AI Content you want to see (Techcrunch). But this very obviously affects only videos that are labeled, right?

📺 Stuff you can watch / read

Got invited to talk TikTok on HIP REPLACEMENT with Kyle Raymond Fitzpatrick (The Trend Report) & Ben Dietz (SIC). We discuss Alt and Elite TikTok Aesthetics back in the days, WarTok and Thirst Trap Propaganda, and The Future of US TikTok and the Mamdani campaign. Thanks so much for having me and my German accent.

What is going on (The Verge)
Deep Lore! (Peripheral Politics)
How to Channel Fandom (Wired)
ICE TikTok (Forbes)
TikTok and Meta in breach of their transparency obligations (EU)

Thanks so much for reading. I am currently coming back to this Verge-article a lot. What is your take on this rock-paper-scissors like take on shitposting and aura farming?

Subscribe now

🦅 U.S. TikTok – ”Great American patriots buying it”

China has not yet approved the deal, additional paperwork is required from both sides, but it seems very likely that TikTok's algorithm will be copied and retrained using US user data as part of a deal to secure the app's future in the U.S. (BBC). Probably in a new standalone U.S. app (FT). And sure, the U.S. government is expected to get a multibillion-dollar fee in the deal (WSJ).

All the final details still need to be nailed down among several joint venture partners that will include Oracle, investment firm Silver Lake Partners and possibly two billionaires — media mogul Rupert Murdoch and personal computer pioneer Michael Dell (AP) – “great American patriots that are buying it — very, very substantial people, people that love our country.”

Bigger picture: We are witnessing an important step in the U.S. administration’s attempt to synchronize the media (“Mediale Gleichschaltung”) – mass media and platforms alike. Both the U.S. and China support ‘info-nationalism’. China has long insisted information flows be controlled domestically, and not by foreign companies or entities. The US has now come to the same conclusion as digital platforms create powerful control points that shape and limit what can be said, read and watched (Al Jazeera).

What will U.S. TikTok’s feeds look like: We do not know yet. Together with Jessica Maddox and Tom Divon I am speculating a bit in this Fast Company piece by Chris Stokel-Walker. As with Twitter/X the future for TikTok in the US has never looked as grim. It might be nice to study, but most likely not so nice to use in private. Unless you are into Rapture TikTok, the dark absurdity of events like the Kirk funeral fest (Garbage Day) or Pokémon theme song based White House Propaganda (TikTok).

🪩 Tiktokification of Politics – Dancing is over

Political communication with the under-30 generation takes place primarily on TikTok and Instagram, according to a new study by the Bertelsmann Foundation and others (FAZ). It is called “How to Sell Democracy Online (Fast)”: Social media (74%) is by far the place where young people obtain political information – ahead of family (58%), school (60%), or friends (54%).

I especially liked this little info snippet in the report (Executive Summary in German): Among young people, dancing in political videos has a significant negative effect (-45%) on the likelihood of wanting to continue watching. This echoes other findings (Big Issue): “A lot of [young people] felt they were cringy. They were forced attempts at humour. And speaking to media and comms people at political parties, many of them didn’t feel necessarily comfortable in producing that kind of content.” I repeat: authenticity, relatability, and added value for your target audience are key!

Someone who has not yet seen dancing is German chancellor Friedrich Merz. He has stepped up his platform performance. While you could actually feel the pain while being recorded throughout the campaign leading up to the German Federal Election 2025 as Jolan Geusen and me observed in our report, his TikTok game has become pretty solid. I discussed with Der Spiegel (german).

📺 TikTok Live is chaos

Then I fell down the rabbit hole that is TikTok Live, writes Natalie Jarvey: “Spend just a few minutes with the feature, which allows anyone to broadcast from their phone with the click of a button, and you’ll feel like you’ve witnessed the full range of the human experience.”

My research on 🫡 TikTok Livestreams (#143) this summer focused on rightwing extremism on the platform as previously discussed. Thanks to a shared effort we can now present an entire report: Live-Streaming: Mapping Networks of Influence and (Dis)information Flow. TikTok Live has become a propaganda infrastructure that operates as an attention economy of hate. We identify a circular mechanism of extremism-as-entertainment, where Nazi propaganda is gamified through play and monetization. Shoutout to the entire fabulous team: Tom Divon توم ديفون, Esther Hammelburg, PhD Pieter van Boheemen, Lea Frühwirth, Jonathan Klüser, Nina Steffen, Laura Postma, Edan Ring, Giulia Costanzo & XINLU WANG.

📟 Synthetic Propaganda

Building up on previous research, i have finally written a paper on synthetic propaganda. It is still under review, but i presented some first observations in a webinar for Disinfo Lab: Synthetic Propaganda – Generative AI and the Future of Political Communication

🗃️ What else

• TikTok Reaches 200 Million EU Users (SocialMediaToday)

• Israel Brings U.S., Israeli Social Media Influencers to Gaza Aid Sites to Fight 'Hamas Hunger Campaign' (Haaretz)

• The young American female soldiers of TikTok (Economist)

• TikTok to replace trust and safety team in Germany with AI and outsourced labor (Guardian)

• This Is the News From TikTok (Atlantic)

Thanks for reading. If you're interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch — I'd be happy to collaborate. Here is Linkedin. Here is Bluesky. And here is a good TikTok.

Subscribe now

📟 The AI propaganda revolution

I introduced the term synthetic propaganda in May (#142) to better understand and unwrap “deliberate, systematic attempt to shape perceptions and influence through GenAI-generated content.” Things have further escalated since i started documenting use cases by Donald Trump. He for instance reshared an AI-generated video of himself playing “Don’t Stop Believin’” by Journey (Source) and posted an image of himself and alligators wearing ICE-caps (Source) just days after Homeland Security had posted the alligators in an attempt to memefy monstrosities on X (Source).

As digital spaces are becoming battlefronts for geopolitical influence (The Conversation) the Iran-Israel war (13 June 2025 – 24 June 2025) led to an array of AI war spam with both conflict parties engaging in an AI slop fight: “AI slop is not the sole domain of anonymous amateur and professional propagandists any longer. The leaders of both Iran and Israel are doing it too. The Supreme Leader of Iran is posting AI-generated missile launches on his X account, a match for similar grotesques on the account of Israel’s Minister of Defense.” (404 media)

As new tools like Google’s Veo 3 make AI-generated videos more realistic than ever, they are seemingly being used to generate videos depicting racist and antisemitic tropes. Media Matters has identified “racist AI-generated content that is explicitly anti-Black, using images of monkeys to promote stereotypes about Black people”. Other racial and ethnic groups were targeted as well. In german context i have been following the #nogaboga-meme (1963 posts) that cloaks anti-Black sentiments in supposedly “funny” content. Can you help me to fully understand the meme?

📆 I am currently gathering data for a paper on synthetic propaganda that i will present at the Annual Conference of the German Society for Media Studies in September.

🫡 TikTok Livestreams

Approximately 130 million creators around the world go live on TikTok each day, reaching billions of viewers (TikTok). Among the streamers are far-right activists, nazi rappers and politicians tying bonds, promoting hate and spreading propaganda. Yet this space has remained largely unresearched due to its ephemeral content, the constant notion of in- and out-groups, missing tools and methods.

Therefore i was super excited to join a team of international researchers in a data sprint during the Digital Methods Summer School at the University of Amsterdam. Thx so much for reaching out Esther Hammelburg. We are currently monitoring more than a hundred German right wing extremists, immersing ourselves in their communication strategies, unwrapping topics, tactics, codes and connections.

👾 Our data relies on an automated monitoring system built by Pieter van Boheemen (Github). If you want to discuss first findings, just drop me a line.

🪱 Malicious Earworms

Together with Marloes Geboers (Google Scholar) i have published a pre-print of a 📘 book chapter on “Malicious earworms and useful memes, how the far-right surfs on TikTok audio trends” (arXiv). It is part of the book “Content Moderation across Social Media Platforms”, edited by Richard Rogers (Google Scholar).

Building up on previous work like the sound of disinformation together with Tom Divon (Google Scholar) we explore the networked linkages between an AfD campaign and extremist rightwing niches on TikTok, ‘held together’ and networked through sounds that hijack popular songs or songs that are blatantly problematic in their contents but that latch onto meme templates to attain visibility.

We presented our results in a keynote for this year’s Digital Methods Summer School at the University of Amsterdam, including lots of audio examples ranging from generative AI techno bangers promoting the AfD, to banned extremist synth pop from the early 1990s to JumpStyle Mashups featuring Sponge Bob.

📆 We will present further sounds at the “Sounds of Fascism: Media, Practices, Imagination” conference at the University Bonn in September.

Tagesschau TikTok

The Tagesschau is probably the most prestigious news service in Germany. It has a pretty solid TikTok account (1.9M Followers 72.4MLikes). I was happy to join a video as TikTok-researcher discussing political communication and the performance of Germany’s new chancellor Friedrich Merz on the platform (“Not that bad, actually”). The video has currently 2.4 M views 🤯.

What else?

TikTok’s research API is broken (AI Forensics). This again leads to a great data standoff. It is researchers vs. platforms (Goanta et al.). While APIs are closing, we are witnessing a new AI (research) rising.

The Reuters Institute Digital News Report 2025 is out. It has 171 pages. TikTok is mentioned 143 times. In 2020 it was mentioned 3 times (#104). Here is the PDF.

Content moderators say they’re exposed to graphic violence, psychological trauma, and union-busting tactics, and now a larger movement is brewing. The people who clean up your TikTok feed are starting to fight back (Rest of World)

Thanks for reading. If you're interested in tailored insights, workshops, consulting, or policy support, feel free to get in touch — I'd be happy to collaborate.

You are reading Understanding TikTok. My name is Marcus. Have we yet connected on Linkedin?

176 years a day. That is not how old i feel after a day with my kids but the amount of videos uploaded to TikTok. A team of scientists from Canada, the US and Germany has succeeded in reverse-engineering unique identifiers to obtain a complete slice of TikTok (Fast Company | Paper). Using this dataset they estimate that 269 million posts are produced each day. Honestly, i am really stressed trying to watch them all.

🍔 The German TikTok Election
📟 Synthetic Propaganda
🕳️ Signs and symbols

Subscribe now

🍔 The German TikTok Election

Together with Jolan Geusen (Linkedin) i have written an analysis of the 2025 federal election campaign on TikTok for Friedrich-Ebert-Stiftung. It just got published here (Website / PDF). We are proud to provide insights on how official and unofficial political accounts were used during the campaign, what strategies proved to be useful or not and we included a set of learnings and policy recommendations.

The 2025 election underscored TikTok’s central role in political campaigning. All major parties engaged actively on the platform, shifting from traditional, rational communication to emotionally resonant, platform-native content.

Especially AfD and Die Linke thrived via unofficial support networks, exploiting participatory affordances. Yet the platform's volatile, trend-driven dynamics demand constant adaptation. To use TikTok effectively for political communication, actors must understand and engage with its unique platform logic. It requires content tailored to its aesthetics, emotional resonance, and trend-driven dynamics. Here are our recommendations:

1. Platform-Specific Content Creation

TikTok is not just a smaller YouTube or an ad channel; it’s a fast-paced, visual, emotional ecosystem. Success relies on producing content that aligns with TikTok’s aesthetics—memes, sounds, vibes. Example: Robert Habeck’s viral “bus driver wave” (Focus) worked because it was tailored to TikTok’s style.

2. Authenticity and Relatability

Politicians gain traction by showing personality and everyday life—not just sharing policy points. Emotional proximity through informal formats (e.g. in a car, at home) builds credibility. Humor, irony, and self-irony resonate strongly on the platform.
Example: Heidi Reichinnek uses direct, unfiltered content to build parasocial relationships and reach (Guardian)

3. Participatory Community Engagement

TikTok thrives on co-creation; one-way communication is ineffective. Parties should enable remixing and adaptation of their content (e.g. via sounds, challenges, downloads). Warning: While the AfD uses intransparent supporter networks, democratic actors should prioritize fairness and transparency – no covert campaigns or bots.

4. Trend Competence and Agility

TikTok requires fast, responsive content production – missed trends mean missed opportunities. Social media teams must be agile, experimental, and platform-literate. "Trial and error" is part of the process; low-effort clips can go viral if they match the TikTok vibe. Example: Friedrich Merz with sunglasses and a Snoop Dogg track –lightweight but trend-aligned (It's the motherfuckin' D-O-double-G).

5. Strategic Influencer Collaboration

Influencers offer built-in trust, reach, and access to niche communities. Effective cooperation requires authenticity and equality—not traditional advertising. Formats like co-creations or challenges work better than scripted promos. Example: Olaf Scholz’s collaboration with Brooklyn (Das Original) showed planned content can succeed if it fits narrative expectations.

Conclusion: TikTok success is not about simply posting — it's about playing the platform. Politicians must be authentic, trend-aware, collaborative, and data-informed to cut through the noise and engage fragmented, digital-native audiences. This may sound obvious, yet rich data lies beyond reach metrics, in understanding how content connects emotionally and socially and can be tailored to user needs.

Link to the entire report here (german). And i am available for briefings, workshops, or strategic exchanges – just drop me a line if you’re interested.

📟 Synthetic Propaganda

There is an endless array of AI-images and videos depicting politicians all around the globe. But there is a special place (Morrissey) reserved for someone special. Images and videos of Donald Trump are disseminated on all platforms, by various actors, including the US president himself, his team and the people responsible for the official White House accounts.

This is interesting because it is unprecedented. As the visual language, the tools used, and the intention (trolling) of this content are basically the same. It does not matter any longer if you are a 13 year old kid with a computer or a well staffed press office when your president wants to mimic the pope or a He-Man meets Star Wars villain. I have started to collect empirical evidence of transgressive content, papers and articles (for a paper) here.

Super happy, that the EU Disinfo lab has chosen my first collection as recommended read – “a must-read for anyone concerned about the role of AI, media manipulation, and disinformation in today’s digital world.” Thanks Ana Romero Vicente!

🕳️ Signs and symbols

"German kids often have little to laugh about, with three or four Germans per 25 students in their classes. German youth are intimidated and often outnumbered; they adapt because they usually have no choice." These lines, sung by a young female voice in German, accompanied by gentle, melancholic music, waft through TikTok (taz).

The sound is used both by people from migrant and left-wing communities and by racists and far-right users. While the latter take the text literally, the former resort to a form of comic relief in a not-so-funny reality of exclusion and everyday racism.

What can be observed here is a new ambivalence in the usage of signs and symbols in certain trends, sounds, brands that carry a variety of meaning depending on the context. Examples include: the brand North Face, being worn by young far-right and far-left activists. The song L’amour toujours by Gigi D’Agostino used in the context of German neo-nazis, paneuropean propaganda and various other political messaging.

I had to think of Pepe the frog (Know Your Meme) and this good article on Hyperallergic: On Twitter, Frogs Were Mostly Nazis, On TikTok, They’re Often Queer. How frogs, more recently seen as a symbol of hate on Twitter, are being queered by TikTok users.

This ambivalence defies fixed attributions and strenghtens uncertainty in an unlimited information environment. Everything is everything (Phoenix). While we are in the midst of a culture clash. Fought (The Clash) on platforms.

📠 What else

TikTok breached digital advertising rules, European Commission says (Politico)
TikTok hit with €530M fine after illegally sending users’ data to China (Politico)
Trump says he’s willing to extend pause on TikTok ban again (The Hill)
The Runaway Tradwives of TikTok (Elle)
TikTok launches TikTok AI Alive (Techcrunch)
Russia’s Use of genAI in Disinformation and Cyber Influence (CRC)
Gendered Harassment During the 2024 EU Parliament Elections on TikTok (ISD)

I have a couple of free slots in the coming months. Happy to offer customized inputs, workshops, consulting or policy papers. If you want to collaborate or are just curious to connect – tap the little letter now 💌

You are reading Understanding TikTok. My name is Marcus. I am a researcher and consultant. Altogether now we are 8K. Thanks for subscribing! Since March 4th (#140) the techno-fascist (New Yorker) coup (Timothy Snyder) lingers on and “absurd and incomprehensible things keep happening” (Albert Lloreta). Oh, and TikTok was not sold. The deadline was extended for another 75 days as the future of the app is now at the center of a geopolitical tussle between Washington and Beijing. The entire thing is likely to become even more complicated as a trade war between the two countries escalates (The New York Times).

Today we talk about:

🤯 Military Brainrot
🇪🇺 Pan-European Propaganda – again
🪇 Some Work Stuff and Links

Subscribe now

🤯 Military Brainrot

Tralalero Tralala. If you have been on TikTok in the last months you have probably come across Italian Brainrot. It refers to a series of brainrot-style AI videos, art and memes that combine animals with objects like planes, coconuts or give them anthropomorphic features, especially legs and feet. An ever weirder variation of German Brainrot (#139).

The videos are typically narrated by a male Italian text-to-speech AI voice rhyming the characters' names with absurd statements about religion or random topics. The memes started in early January 2025 on TikTok with Tralalero Tralala, a shark wearing Nike shoes (Know Your Meme).

As always the “trend” was incorporated by brands, political parties and – in recent times – by actors of the military-industrial complex. Just like other “trends” with the IDF utilizing the Studio Ghibli AI Generator (Know Your Meme) further pushing the limits of military cutefication or the White House where someone embraced ASMR (Know Your Meme) turning government communication further into “gleeful cruelty” (The Atlantic).

Airbus is Europe's largest aerospace and (after BAE Systems) second largest defence company. Airbus has been using trending sounds to successfully promote the company on TikTok. The account (@airbus) has garnered 3.4 M likes. One of the videos shows an A400M-edit. We see video snippets of the Airbus A400M Atlas, a European four-engine turboprop military transport aircraft, edited to Doechii’s viral ‘Anxiety’ song (indy100) – a rather happy song about anxiety disorder.

In another video we see footage of the Eurofighter – a supersonic, canard delta wing, multirole fighter. The video is edited to the Italian Brainrot sound Ballerina Cappuccina – a young woman with a cup of coffee replacing her head who specializes in Ballet. She was Capuccino Assasino's wife, but got divorced after she was brutally cheated on (Brainrot Fandom).

The caption of the Airbus video: Our kind of ballerina 🩰 #Eurofighter #airbus. The Brainrot-reference totally makes sense given the fact that the entire trend gained viral traction with Bombardiro Crocodilo, a military bomber plane with a crocodile face. The male Italian text-to-speech voice here says, "Bombardino Crocodilo, an alligator that bombs kids in Gaza and Palestine and doesn’t believe in Allah and loves bombs."

“We no longer understand the reality around us and live in a state of growing confusion”, writes Albert Lloreta: “It seems that the system of values and certainties in which we grew up is collapsing, but we go on with inertia. We cling to the fiction of normality in our lives and pretend not to see how the strange stain in the sky grows day by day.” Tralalero Tralala.

🇪🇺 Pan-European Propaganda – again

I am still not sure what to make of this. This is how i concluded first observations of Pan-European Propaganda on TikTok in March (#140). Thankfully Szilvi Német (x), a Budapest based journalist, art historian and media researcher took over, did a decent investigation and published this brilliant deep dive. We wrote back and forth a bit, here are some questions and some answers on the topic:

What surprised you the most? And why?

I’ve never seen — or imagined — the European Union like this. Especially in Hungary, there’s a strong tendency to depict the machinery of Brussels as something dull, grey, and overly bureaucratic. There’s even an unspoken rule in journalism: don’t put “EU” in the headline, because no one will click on it. Seeing the union reimagined as a military superpower — clad in bold, militant, high-tech visuals — opened up an entirely new dimension for me.

While working on the article, I reached out to many creators, and what really struck me was how ordinary people are investing time and effort into promoting ideas they deeply believe in — even setting aside nationalist sentiments in favor of a shared European cause. They genuinely want to take pride in being European. To me, that feels like a historic moment.

What are typical audiovisual ingredients of the videos?

My friend and I have this running joke that L’Amour Toujours by Gigi D’Agostino is the unofficial anthem of this European wave. There’s another propaganda track called Make Europe Great Again — an upbeat, anthemic march with strikingly powerful lyrics. It’s really incredibly catchy. After I heard it for the first time, I couldn’t get it out of my head for days. But it's also everything from MGMT’s Little Dark Age to Taylor Swift’s …Ready for It?

Visually, the content tends to fall into two main categories. One is the remake or appropriation of 20th-century agitprop posters, reimagined with the help of AI. Using Midjourney, you can even see these static posters come to life — jets literally fly off the two-dimensional surface.

The other category is entirely new AI-generated imagery. Some of it is made with older AI engines, giving it a retrofuturistic aesthetic, while others are hyperrealistic — almost veering into kitschy, Disney-like territory. Stylistically, the deep blue of the EU flag serves as a unifying thread, tying together this wide variety of images.

Thematically, the building blocks are strikingly consistent: the mythical figure of Europa Invicta, iconic European architecture, the act of planting the EU flag on the Moon, high-tech European trains and rockets, and — most importantly — a unified European army, often portrayed as a kind of modern-day Noah’s Ark featuring a mix of ethnic phenotypes representing different nations. From there, the narrative branches out. Some accounts promote nuclear energy, or the support of Ukraine, while others issue practical calls to action — like urging followers to buy exclusively European-made products.

What is the most interesting from an aesthetic point of view?

There’s been a lot of discussion about how AI-generated imagery seems to naturally align with far-right aesthetics — and I see echoes of that in the pan-European visuals as well. In this trend, nationalism and patriotism appear to shift from individual countries to the continent as a whole. Beyond the glorification of military culture and warfare, these images often project a strong sense of superiority — a kind of continental pride with an unmistakable edge. Others seem to sense that too, as many commenters pointed out the sinister undertones. Users referred to it as “eurofash” (short for euro fascism), evoking uncomfortable echoes of 20th-century authoritarian regimes. They also added that this is definitely not the Europe they want to live in.

How has this “trend” developed?

Many accounts were newly created in late February and early March, following Trump’s announcements and the EU’s rearmament plans. While federalists had been promoting similar ideas long before, now they seemed to have gained renewed momentum and visibility.I’d say it’s still a microtrend — some videos can rack up as many as 100,000 views, but the accounts behind them haven’t yet built a critical mass of followers.

Among the subgenres, there's a distinctly European version of MAGA emerging, rooted in the belief that we're living in a new imperial age. In the U.S., some commentators have described Donald Trump as a modern-day Caesar — and similar narratives are now being adapted to the European context. You see this when figures like Emmanuel Macron or Ursula von der Leyen are depicted as Roman emperors, dressed in togas and crowned with laurel wreaths.

Who is behind it and what is the goal here?

Most accounts are run by individuals who support the idea of a more federal, united Europe — one that can stand up to the Russian threat and act independently of an unpredictable United States. Alongside them are a number of channels operated by small pan-European organizations and media platforms, such as the Pan-European Foundation and Forum Europaeum. The content is largely fan art, often remixed with real speeches from European politicians and references to rearmament plans.

Thanks Szilvi!

🪇 What else – Work edition:

I wrote about Palestinians being falsely accused of engaging in "Pallywood" in November 2023 (#113). Therefore i was happy to be cited in a story by NGO Forbidden Stories – an international network of journalists. New research of Forbidden Stories and its partners have uncovered research from an Israeli NGO indicating that the Israeli X account called Gazawood does practice online disinformation on a massive scale.

VSquare is one of the leading regional voices of investigative journalism in Central Europe. I discussed the German TikTok elections with Josef Šlerka. He wrote a great summary on how TikTok played a crucial role in the election especially for younger generations.

I discussed “Facts, fakes and filter bubbles” with Thomas Laschyk, founder of Volksverpetzer, and Alice Echtermann from Correctiv-Faktencheck, pleasantly moderated by journalist Damla Hekimoğlu for Koerber Foundation in Hamburg recently. You can watch the entire discussion where i try to debunk the whole notion of “filter bubbles” (Video where we all speak german).

In May, I will be teaching two block seminars: "Political Communication on TikTok" at the University of Düsseldorf and "Vertical Storytelling" at the Institute of Journalism at the Hannover University of Applied Sciences and Arts.

More

TikTok is adding Footnotes (Engadget)
Views are nothing. Views are lies. (The Verge)
Lorde (Hypebeast)
Zahide Won This Trend (Know your Meme)
The social media domination of the German far right (ECDA)
News #foryou on TikTok: A Digital Methods-Based Study (Journalism & Mass Communication Quarterly)

Thanks for reading. If you want to discuss TikTok, need consulting, feedback on your account or want a keynote. Come say ‘Hi’.

You are reading Understanding TikTok. It is me, Marcus (Bluesky).The last couple of days have been coined by an absurd AI-video and a not yet seen diplomatic meltdown in the White House resulting in the conclusion, that the current US administration rather sides with Russia than with long-time democratic allies in Europe. Meanwhile on TikTok i have recognized a surge of European slopaganda, as Michał Klincewicz, Mark Alfano, and Amir Ebrahimi Fard have called “The interaction between propaganda and generative AI”. Let’s talk about:

🤖 AI’s New Political Aesthetic: Slopaganda
🗳️ The German 2025 Elections - part 2
📚 Systematic Literature Review

Subscribe now

🤖 AI’s New Political Aesthetic: Slopaganda
From Trump Gaza to a revival of Pan-European TikTok propaganda

By now you all have most likely seen the unhinged video president Trump posted on his Truth Social account, depicting a war-torn Gaza morphing into a Trump-branded beach resort (Garbage Day). It then soon turned out that the creators of the AI-generated video saw it as some sort of “satire” never meant to become a “propaganda machine” (NBC).

Interestingly enough Gemini knew?! For the first time i had used Vertex AI / Google Gemini for some AI-assistance to analyse the video before the origin became known. Gemini’s possible interpretations pointed out: satire of neo-colonialism, critique of celebrity power, commentary on economic disparity and exaggeration and absurdity (see results) as well as pretty useful answers concerning visual style references (see results). Still experimenting with AI-assistance (compare #139).

The video has yet again sparked a discussion wether “AI is *the* aesthetic of fascism” (compare New Socialist). As Roland Meyer (Bluesky) points out: As a nostalgia machine and cliché amplifier, it {genAI} is almost necessary for the creation of right-wing worldviews (Geschichte der Gegenwart).

A couple of days later on TikTok i got confronted with a bunch of accounts posting pan-european genAI propaganda (Bluesky). I am still not sure what to make of this. An aesthetic to prepare for the new geostrategic role of Europe? The necessity has become crystal clear in recent days but who pushes these videos to whom?

The platform and its users have always been a perfect seismograph for shifts in societal moods and vibes in times of crisis. Yet between Eurotok – a mostly friendly transnational discourse (“seems softer, quieter and more interesting and fun” example) – and European Edits – including nationalistic, nostalgia amplifiers – the generative AI powered Pan-European propaganda promotes euro-patriotism with retrofuturistic genAI slop and military beautification. Yet the flag all protagonists wave is a European one, supposed to represent the values of democracy, freedom, equality, rule of law, human dignity, and human rights.

🗳️ The German 2025 Elections - part 2
It is the fan accounts and TikTok affordances: memes, vibes and edits

More analysis of the German Federal elections on TikTok are coming in (compare #139). The Institute of German Economy has published a first report (in german). They coded all videos uploaded by members of the German parliament with a TikTok account by topic. Not sure if a category “personal & memes” really makes sense as i understand memes more as a crucial TikTok affordance for the successful dissemination of basically all topics.

Belltower by the Amadeu Antonio Foundation has published a good analysis (in german) with a bunch of visualisations too. Most striking to me is “the large difference in the number of “fan accounts”. While we registered between 74 (BSW) and 873 (SPD) such accounts for the democratic parties, there were 12,790 for the AfD in the same period.”

These observations are echoed in my preliminary results comparing e.g. #roberthabeckedit (188 posts) and #aliceweideledit (2080 posts). Fans are the core amplifiers using sound memes (“Deutschland braucht die AfD”) and edits in a participatory propaganda approach i am currently investigating.

📚 Systematic Literature Review
A systematic study of political communication on TikTok

Julian Hohner is a political scientist and informatics and “ a hopefully anytime soon to be finished PhD at LMU Munich working with Prof. Dr. Diana Rieger”. His main work deals with radicalisation and extremism research and the role the internet plays in these dynamics but he has just conducted a systematic literature review on TikTok & political communication.

What are your three most important findings?

The TikTok algorithm is really powerful in pushing users into rabbit holes about specific topics or attitudes – probably the most effective in cross-platform perspective.

In a plethora of studies, TikTok shows significant influence on political aspects of our world, positive and negative alike. TikTok use increases the contact and participation in politics and supports community building and expression, but is also associated with a range of problematic behaviour such as radical protest participation, increased likeliness or building far-right attitudes .

In extension, many studies warn of TikTok’s role in mainstreaming, normalising or ‘de-demonising’ radical and extreme narratives serving as platform for far-right (electoral) success.

What surprised you the most when doing your lit review?

The lack of diversity in research about TikTok’s role. Most papers either look at popularity and engagement metrics of specific themes, narratives or content or investigate the recommendation algorithm. Media effects studies, method papers that deal with how to process videos, or studies that focus on unique traits of TikTok such as livestreams, donation system, reposts or community building are rare.

Where do you see the most crucial research gap?

Longitudinal analysis that show possible time affects trends are desperately needed – only four studies could be classified as such. Only those analytical designs can, for example, sufficiently track possible rabbit hole effects or the mainstreaming of ideological extreme narratives. Data donations studies and computational method approaches are possible technical solutions for this.

Can you recommend 1 or 2 „must read“ papers on TikTok & political communication?

Already a bit outdated in this fast-changing environment TikTok is settled in, but the article does a good job in catching the essence of why and what makes political content on TikTok so unique and successful: Cervi, L., & Divon, T. (2023). Playful Activism: Memetic Performances of Palestinian Resistance in TikTok #Challenges. Social Media + Society, 9(1). https://doi.org/10.1177/205630512311576072.

Theoretical article about the danger of normalising extreme narratives and not exactly only about TikTok (but rather on visual heavy platforms): Bonansinga, D. (2024). Visual de-demonisation: A new era of radical right mainstreaming. The British Journal of Politics and International Relations, 13691481241259384.

Thx Julian!

🥐 More

The TikTok ban had almost nothing to do with China but Israel (Taylor Lorenz)
TikTok has NOT revealed its algorithm points system for how it ranks (Linkedin)
How 12 Americans See Life After Watching a Lot of TikTok (NYT)
TikTok restructures trust and safety team, lays off staff in unit, sources say (Reuters)
Half-way point of the 75-day extension TikTok was afforded (Digiday)
2013 core (TikTok)

That’s it for now. Thx for reading. And now please send me your most precious TikTok video. You never do. Here. Thx.

Subscribe now

🗳️ The German 2025 Elections

The defining sound meme on German TikTok right before the federal election has been a nostalgia-infused “we meet again years after a break-up and have a coffee” pop song performed by Florian Silbereisen and Helene Fischer who dated in the noughties. A perfect metaphor for a general “take me back to when everything was somewhat okay” vibe even though it never was – just in time of geopolitical upheaval. Nearly every party has used the song in their campaigns with Olaf Scholz and team using an Anti-Afd parody version of the song one day before the election on his official account.

First take-aways on election eve:

• TikTok for political campaigning has gone full mainstream in 2024/25

• TikTok’s cultural impact is undefeated with an older target audience growing

• Political campaigning in Germany has never been so meme-infused

• Political messaging from posters to platforms is heavily fragmented

• Politicians and parties finally and fully embrace the platform era

• The role of sound memes needs further investigation. Hook me up!

Despite TikTok’s assertions to “protect election integrity” (stop me if you think that you’ve heard this one before) problems already pointed out after Germany’s last federal election in 2021 (Mozilla) still exist. Several studies point out:

• "TikTok displayed the greatest right-wing skew, per its findings — showing right-leaning content 74% of the time." (Global Witness Study, Techcrunch, Nieman Lab)

• “TikTok classifies only 59% of the content of party or candidate accounts and only 47% of the content of fan pages as political.” (ISD Germany)

I am currently scraping data for a more in-depth report on The German TikTok Elections 2025 that will be published by Friedrich Ebert Foundation in the coming months. Wish me luck on my deep dive where i will investigate stuff like this or this this or this.

🤖 Using ChatGPT for analysing TikTok data

Analysing the complexities of multimodal, layered and dense video content poses challenge to researchers. Large language models like ChatGPT might be of help here. But how? I asked Carlo De Gaetano (IT) – designer and researcher with the Visual Methodologies Collective at the Amsterdam University of Applied Sciences.

Carlo presented his experiments in a workshop at a winter school on AI Methods at Siegen University (slide deck) in February 2025 that i had the pleasure to attend. This is the condensed version of a longer Q&A with Carlo.

Hi Carlo, how are you using ChatGPT to analyze TikTok data?
One of my recent experiments focused on maps in sea level rise videos—a recurring visual trope on TikTok. I wanted to test whether ChatGPT could recognize place names, countries, and regions just by looking at maps. After processing around 30 thumbnails, I asked ChatGPT to compile the identified geographical entities into a structured dataset. I was pretty excited by the results. Read more in the full interview here.

How can ChatGPT be most helpful here?
For me, ChatGPT is most useful as a thinking partner—it helps structure and refine my research process, making it easier to go from a messy collection of videos to a more organized way of analyzing them. One of the simplest but most practical ways ChatGPT helps is with search queries. Finding the right videos on TikTok isn’t always straightforward, and hashtags or captions don’t always reflect the full content of a video. ChatGPT helps me brainstorm in the early stages of research to find precise keywords and hashtag combinations that improve my chances of pulling in the most relevant material.

Once I have a dataset, ChatGPT assists in developing workflows for analyzing the collection. As I have already mentioned, since I can’t upload videos directly (for now), I have to rely on extracted frames, video thumbnails, or timeline snapshots to represent the content visually. ChatGPT can help sketch out different ways to process these images systematically: for example, the suggestion to track color transitions over time to detect sea level rise maps came from ChatGPT, and was then refined collaboratively through iterations.

What surprised you the most when analyzing TikTok data with ChatGPT?
One of the biggest surprises was how well ChatGPT remembers past conversations and can build on them. When working on a research project that spans multiple sessions, it’s easy to lose track of earlier ideas or minor refinements that were made along the way. But ChatGPT often recalls previous discussions, references them, and even reminds me of things I had considered before but didn’t follow up on. This makes it feel more like an actual research collaborator rather than just a tool for answering isolated questions. Another unexpected aspect is that ChatGPT doesn’t just give answers—it helps refine questions. When researching something as visually complex as audiovisual collections, it’s not always clear what the best approach is. Instead of just providing pre-formed insights, ChatGPT acts as a sounding board, helping me clarify what I’m looking for and suggesting different ways to structure an analysis.

Also, I was surprised by how much creativity is required when working with AI in this way. Since ChatGPT can’t directly analyze video, a lot of the process involves figuring out workarounds—how to extract meaningful data from videos in a way that AI can process. As an information designer, this has made me think more deeply about how research is structured and has pushed me to develop new methods I wouldn’t have considered otherwise.

What are common mistakes when using ChatGPT for TikTok analysis?
I can share some of my own mistakes while using ChatGPT as a visual assistant. The first one was trying to ask for everything in a single prompt. Another mistake is not refining prompts enough. A vague question leads to a vague answer. For example, when I first started asking ChatGPT to describe TikTok thumbnails, I wasn’t specific enough about the kind of details I wanted. This meant I got overly generic descriptions that didn’t really help my research. But after a few rounds of tweaking, I realized that giving explicit instructions on what to focus on—colors, text overlays, recognizable landmarks—made a huge difference.

I learned the hard way that uploading too many images at once makes things go sideways. At one point, I tried uploading a zip file containing 200 TikTok thumbnails, expecting ChatGPT to neatly categorize them. Instead, I found myself in a strange situation where ChatGPT confidently made up descriptions that didn’t match the images at all—a perfect example of how LLMs, as Harrison (2023) puts it, are "mansplaining machines: often wrong, yet always certain." It was frustrating but also a good reminder that AI models don’t “see” images in the way we do—they process metadata, pixel data, and patterns, but they can still hallucinate descriptions when overloaded. So, the lesson? Take it step by step, test small batches first, and don’t assume the AI is always right—because sometimes, it’s just confidently wrong.

Would you mind sharing 1 or 2 tricks or tips when analyzing TikTok data with ChatGPT?
There is a misconception that because ChatGPT is presented to us as a machine, it must be great at math—which is absolutely not the case. I quickly learned that asking it to count how many times a certain object appeared in a timeline led to nonsense results. It’s much better at what it’s actually designed for—describing images, summarizing text, and categorizing things using a few examples. A useful strategy when analyzing videos is to build a structured tagging system: when working with a collection of TikTok videos, it helps to design a clear system for categorizing images. Instead of expecting ChatGPT to "figure out" a pattern on its own, I use the few-shot learning method, where I manually tag a small set of images with labels first and then have ChatGPT apply those categories to the rest of the dataset.

Another key strategy is to combine AI with other tools to scale up the research. As a visual research assistant, ChatGPT works best when it’s used in combination with other software rather than as a standalone solution.In this way, ChatGPT does not do the heavy lifting itself but helps outline the logic and structure of the solution, while Python hands the execution. AI doesn’t replace manual work, but it can massively speed up and refine the process when used strategically.

Thanks so much!

🍏 What else

There is a new TikTok-Scraper out there (Github)
What we know about TikTok content creators (Pew Research)
Politicisation of influencers workshop coming up (Utrecht University)
Why TikTok’s loss might be Instagram’s gain (DigiDay)
Wie beeinflusst TikTok die Wahlentscheidung? (BR)

That’s it for now. I am in Innsbruck next week for the Annual conference (Jahrestagung) 2025 of the specialist groups “Communication and Politics” (DGPuK), “AK Politics and Communication” (DVPW) and “Political Communication” (SGKM) to present on TikTok & participatory propaganda in the context of elections campaigns. Are you nearby? Come on over. Speak soon. Ciao

P.S. Anyone else’s FYP escalating between Snailrave and Party 4U?

You are reading Understanding TikTok. It is me, Marcus. Germany is electing on February 23 (Politico). I am checking political vibes on German TikTok these days. Parties, politicians and supporters have started to fully embrace TikTok with still-chancellor Olaf Scholz and team going fully ballistic –– within the scope of his possibilities.

❤️ ReclaimTikTok: Mission accomplished
💙 Synthetic Propaganda: AfD Slop

Subscribe now

❤️ Mission accomplished: TikTok reclaimed √

After extensive media coverage concerning the dominance of Germany’s right-wing extremist party AfD on TikTok (#123) several attempts of countermovements could be observed in the German part of TikTok (#126) throughout 2024, most revolving around #ReclaimTikTok (70.7K post).

While hashtag activism already felt a bit dated in 2024 with hashtags losing their relevance on the platform (TikTok Audit) democratic creators opposing the AfD have demonstrated a steep learning curve concerning platform specifics and meme-driven amplifying strategies in recent months. These include active trolling, mocking and comment bombing. Trending sounds, memes and vernaculars are produced, adapted and amplified. Here are some recent examples of political communication on TikTok beyond traditional and or official accounts.

🧠 German Brainrot

Between November 2024 and January 2025 videos featuring rendered animals with German language voiceover gained traction on the platform (Know Your Meme). A video featuring a rendered pigeon inside a supermarket buying daily items while being narrated by an AI-generated German voice received roughly 9 million plays and 1 million likes in two months. Many variations of this absurdist Germanrot content – referencing brain rot (Wikipedia) – occurred. Aidan Walker going more into detail on Germanrot here (Video 4:42 min).

A new sort of politically charged content including rendered animals could be observed in the last couple of days. On February 3 user @germanbrain2 uploaded a first video on his account. It includes a pigeon in a bathroom discussing AfD chancellor candidate Alice Weidel, her female partner and – according to the pigeon – a perceived inconsistency when contrasting her lifestyle to the political program her party stand for.

The video has garnered 220.7K views in three days. The account demonstrates a successful combination of a temporary meme and political content suited for the specific platform and its target audience.

🎷 Remix featuring Glööööckler

Stitches and duets including AfD personal and creators that were meant to debunk or oppose the party have been quite common on the platform. Yet many of these lacked a particular entertaining aspect.

Some days ago the account @glockeler.gegen.rechts started posting videos. These include stolen video snippets from a fashion reality competition series featuring German fashion designer and entrepreneur Harald Glööckler, well known for – well – basically his face. He underwent numerous cosmetic surgeries to change his face according to his own ideas. Snippets of Glööckler (who is well-known for his temper) are edited in order to appear as if he directly reacts to several AfD politicians criticizing them. The editing is well executed, the content is entertaining and delivers enough hooks to keep users engaged even after the initial “shit, look at that”-moment.

🥁 Deutsche Sound Propaganda

TikTok is a sound-driven platform. Sound is its backbone. German politics has embraced that, examples range from Pro-AfD folk songs to leftist techno-remixes of German politician Gregor Gysi (84.3K Likes). Interestingly though more and more users not only re-use given sounds but compose political protest songs.

User @KeMax set up an account November 22, 2024. Since then he has published six videos with songs criticising Friedrich Merz, chancellor candidate for the Christian Democratic Union of Germany. Until now his songs only reach a very limited amount of users (1,591 Likes), yet he stands pars pro toto for a new breed of politicized musicians.

All these examples (there are way more) show creative and platform-specific forms of political communication outside given structures and institutions. They demonstrate how TikTok serves as an important third space for political discourse that keeps evolving rapidly.

💙 AfD Slop

Since Hurricane Milton in 2024 we are in the midst of the the ‘fuck it’ era of AI slop (Jason Koebler) where a lot of people actually do not care if images are “real” or not as long as they feel real. I have written about AI Slop in October 2024 (#135) but only recently has my For You Page been infused with more generative AI videos promoting the German far-right party AfD, their worldwide and aesthetics.

Several state associations of the AfD have used generative AI to produce campaign videos or posters as a fast and convenient way to create tailored content. AI-generated clips here serve as nostalgia machines and emotional cliche amplifiers with imagery referencing historic aesthetics from the 19th and 20th century. Roland Meyer (Bluesky) has written about that (Geschichte der Gegenwart).

Besides these official examples, supporters produce variations of low quality media using generative artificial intelligence to help amplify pro-AfD messages. These sometimes bizarre AI fan productions (think of blond kids with blue AfD dresses riding on Doberman dogs going backwards in a parade of German flag waving adults) capture attention.

But they do more: the imagery combines elements of nationalism, strength, clear enemies and emotional appeal to promote the AFD as a protector of Germany’s heritage and future. The use of children and animals humanizes the party's image. By embedding these elements in celebratory and positive contexts, the images aim to soften the AFD’s controversial reputation and normalize its ideology.

This kind of tailoring can not be achieved with normal stock images. Instead it creates a unique *hyperreal aesthetic that feels uncanny, simplistic, meme-infused and very platform-specific. Here is a good article on how to read an AI image: Salvaggio, Eryk. 2023. How to Read an AI Image: Toward a Media Studies Methodology for the Analysis of Synthetic Images. In: IMAGE. Generative Imagery: Towards a ‘New Paradigm’ of Machine Learning-Based Image Production, 19(1): 83–99. DOI: https://doi.org/10.25969/mediarep/22328

* Hyperreal is seen as a condition in which what is generally regarded as real and what is understood as fiction are seamlessly blended together in experiences so that there is no longer any clear distinction between where one ends and the other begins

🤍 What else?

The TikTok “ban” as fiction. How legal and political symbolism created the “ban” narrative (Verfassungsblog)

Doug’s winter party (Dexerto)

TikTok Stole Our Hearts, but Can It Last? (Wired)

TikTok User Age, Gender, & Demographics (Exploding Topics)

TikTok-Zahlen zur Bundestagswahl 2025, KW 05 (German)

Thanks for reading. Speak soon. Ciao, Marcus

I have seen many people crying on the app that still is an essential transnational third space (default.blog) even when the vibes are off (Washington Post) and the golden age (Dazed) is over. It is still too early for a definite obituary (The Hollywood Reporter) as we will see more mayhem, more zigzagging, and more memes as we are living through a coup d'état, pretty wild (G. Willow Willson) a little perplexed (Justin Trudeau) observing mad king behaviour (Aaron Reichlin-Melnick).

🤡 From Hopecore to Hopelesscore
🦅 2025 German federal TikTok election
🛕 Academia: New TikTok papers
🪭 What else

Subscribe now

🤡 From Hopecore to Hopelesscore

The time span between #136 and now coincidentally fits a memetic cycle that Adam Aleksic perfectly describes in his newsletter The Etymology Nerd: The meme genre (hopelesscore) — in which depressive language is presented motivationally to mock the earlier “hopecore” aesthetic — first started taking off on TikTok in November 2024.

Yet again memes as the lingua franca of the internet provide a sensitive seismograph for societal, political, mental shifts. Back in November 2024 German chancellor Olaf Scholz and his TikTok team for instance used a hopecore-video (2024-12-11) to communicate the end of his coalition (Spiegel). A somewhat dark, macabre Gen Z take anticipating hoplesscore as the remaing governing coalition is nowhere near a sufficient majority to win the early elections on February 23.

The memetic development from hope to hopeless and the following containment turning the trend into mere “‘content’ like any other online aesthetic, resigned to being farmed by TikTok spam accounts” (Aleksic) is a perfect metaphor for the current status quo of things. On to new meme cycles.

🦅 2025 German federal TikTok election

A federal election will be held in Germany on 23 February 2025 to elect the 630 members of the 21st Bundestag. Originally scheduled for 28 September 2025, the elections were brought forward due to the collapse of the governing coalition during the 2024 German government crisis.

All parties represented in the Bundestag rely on TikTok in this election cycle to reach 22.8 million TikTok users (population 84.5 million). While the last election in 2021 saw German political parties taking a more traditional sender-receiver approach, despite the far-right AfD, this time all parties take a more platform-literate approach, utilizing memes, trending sounds and outsourcing the distribution to authentic and inauthentic fan accounts.

I am currently collecting data for a report, but i will share some first observations in the coming weeks. In the meantime, here is something i wrote for the Mozilla Foundation in 2021: Broken Promises: TikTok and the German Election. And here is one on the last European elections 2024 for Friedrich Ebert Stiftung, german only sorry: Von Reichweite und Algorithmen: Analyse des Europawahlkampfs ausgewählter Parteien auf TikTok.

🎤 Quote

I think we’ll look back on these five years as an inflection point for social media, the era when the internet officially became real life
Rebecca Jennings

🛕 Academia: New TikTok papers

Fichman, P., & Akter, S. (2025). Political trolling on TikTok. Telematics and Informatics, 96, 102226.

Kalnes, Ø., & Bjørge, N. M. (2025). ‘So, we have occupied TikTok’: Ukrainian women in# ParticipativeWar. Media, War & Conflict, 17506352241307010.

Tanner, S., & Gillardin, F. (2025). Toxic Communication on TikTok: Sigma Masculinities and Gendered Disinformation. Social Media+ Society, 11(1), 20563051251313844.

Ibrahim, H., Jang, H. D., Aldahoul, N., Kaufman, A. R., Rahwan, T., & Zaki, Y. (2025). TikTok's recommendations skewed towards Republican content during the 2024 US presidential race. arXiv preprint arXiv:2501.17831.

Reinis, S., & Laughlin, C. (2025). “GOD IS MY SPONSORED AD!! MY ALGORITHM!”: The spiritual algorithmic imaginary and Christian TikTok. New Media & Society, 14614448251315128.

🪭 What else

TikTok's 'cute winter boots' trend explained (Usermag, 25 January 2025)

“TikTok refugees” drove a global Xiaohongshu boom (Rest of World, 24 January 2025) Here is an academic reading list

‘There Was Nothing Ever Unique About TikTok And I Can Prove It (Garbage Day, 20 January 2025)

Local news has invented another fake TikTok challenge (Usermag, 10 January 2025)

There is more to a political earthquake then TikTok in Romania (Eastblog, 19 December 2024

The year in brainrot (New Yorker, 16 December 2024) Is brainrot a myth (Dazed, 29 January 2025)

🎻 What i have been doing

Finally finished a version of a paper together with Tom Divon: Dancing in Uniform: The Role of Thirst Trap Propaganda in Military Image Wars on TikTok. I presented it at AoIR 2024 in Sheffield, Axel Bruns summed it up nicely here. Guest lectured on the same topic at University of Alberta, Media and Technology Studies. Thanks for having me Nicolette Little.

Joined Europe’s first Political Tech Summit to talk about Memes as Modern Soft-Propaganda-Tools and started a Systematic Literature Review on Memes of Misinformation.

Wrote a little essay on Postdigital Propaganda that got instantly desk rejected.

Taught two courses on TikTok for political communication at TH Köln – University of Applied Sciences, focusing on TikTok campaigns ahead of Germany’s Federal Election 2025. Will teach another course on political TikTok campaigns for a Master’s program in Political communication at Heinrich Heine University Düsseldorf in the coming semester.

Evaluated some journalistic TikTok formats, presented at Fraunhofer Gesellschaft, talked about Germany’s far-right on TikTok and joined a symposium on “Escalation”.

Thanks for reading. All typos are mine. Speak soon. Best, Marcus

Subscribe now

🔵 Team Harris

“Team Harris outraised, outspent, out-advertised, and (for the most part) out-posted Trump across the internet. Their senior staff understood the changing media environment and how social media platforms have evolved since 2020 - and they adapted their strategies accordingly. Still, it wasn’t enough.” Kyle Tharp, FWIW

The digital rapid reponse team – The ‘feral 25-year-olds’ making Kamala Harris go viral on TikTok (Washington Post) – has done nothing particularly wrong. Harris soundly won the vote among 18- to 29-year-olds. But TikTok as all other platforms is only part of a joint effort. Future elections will not be won without an online ecosystem. The internet is real life now,  tweets Parker Butler, director of digital rapid response: Democrats need to drastically grow our online ecosystems to reach audiences far outside of the echo chambers.

Adam Aleksic points to a structural aspect (Usermag) – broadcast vs. viral communication – and concludes: Candidates are reshaping their communication to perform better on algorithmic platforms: getting more likes, more shares, and more retention—which means exaggerating and hyperbolizing, since less exciting content simply doesn’t go far on those platforms. Therefore: All political campaigning is going to be meme-based in the future, because algorithms are only going to tighten their grip over our society.

🔴 Team Trump

It would be pointless and barely possible to list all of Trump’s vulgarities during this campaign. Yet he mastered the game of attention with a miraculous agility by stimulating – to pick out only one striking example – oral sex on  a microphone (Independent). Well, politicians are engaging in greater online incivility than before because they know that civil content garners less engagement (Frimer et al.).

Trump has turned into a living meme. With or without irony people out there mimic his moves (Learn the Trump Dance), turn his weird ramblings into songs (Eating the dogs) or wear bandages on ears in support (The Guardian). Trump’s pop cultural appeal throughout the years (Home alone etc.) has entered its most shrillest phase with Trump Latinos rapping for the convicted felon while praising “Black Maga” (@TrumpLations24) and religious AI slop edits where he can be seen standing next to Jesus both holding a US flag. 

And then there are friendly accounts like @americafirst617, @realmichaelsolakiewicz, and others that would show behind-the-scenes footage of Trump interacting with his grandkids or joking with supporters on the golf course, often receiving millions of views. That side of Trump had rarely been shown in the mainstream media or on other social media platforms as Kyle Tharp, FWIW points out. Meme-infused content, perfectly prepared for participatory propaganda where the dissemination of content is outsourced.

Besides all that, the Trump campaign perfectly aligned to a broader “cultural shift {that} has been brewing for years in America as Paul Anleitner points out on Twitter. While playful postmodernists vibed with an “Eating the dogs” edit, a lot of others condemned the already debunked “fact” of immigrants eating pets in a form of
post-ironic sincerity.  

⚪ Six Thoughts on Political Campaigning

Political campaigning must embrace the online-offline nexus. The internet is the real life. And the real life is the internet.

Political campaigning needs to reach very different target audiences on all platforms available. Therefore political campaigning needs an online ecosystem. To amplify your message you need people willing to help spread it. 

Emotion > content. There has been lots of talking about vibe and vibe shifts (NYT). Make people “feel their way into politics” in ways that evolve beyond the conventional, deliberative logic of traditional public sphere (Papacharissi 2014)

Political campaigning needs to embrace memes as lingua franca of the internet (Galip 2023). Memes are not merely funny images with text but relevant carriers of meaning. They help in building communities. They help in spreading ideas. They help in engaging others. They help making ideas stick.

Political campaigning has to fit in national / regional peculiarities. Ideas can not just be copied. Parts of the German Greens for instance still try to adopt the entire brat summer despite the fact that it is officially over. 

Political campaigning needs to be fast and ferocious. In order to get people in an infinite media environment. In order to simply keep up with the maturity of information campaigns by bad actors. 

🟡 Quote 

One exercise that I think would be useful for understanding the effect of TikTok on the electorate would be to step back and think less about the type of content that you see there and think more about the type of person that TikTok shapes its users into being.

Max Read, The TikTok Electorate

🟠 What else 

⚬ The German Foreign Office is on TikTok now (@aussenpolitik)

⚬ Four months of campaigning, conspiracy myths and dancing (Die Zeit)

⚬ Reading the Auras of Democrats and Republicans on the Eve of the Election (NYT)

⚬ Visual Politics: TikTok versus Instagram in the 2024 U.S. Presidential Election (Preprint)

⚬ Beyond Anti-Elitism and Out-Group Attacks: How Concerns Shape the AfD’s Populist Representation on German TikTok During the 2024 European Elections (Preprint)

⚬ Why is the 4B movement going viral after Trump’s win (Guardian)

⚬ Trudeau government bans TikTok from operating in Canada — but Canadians can still use it (CBC)

Thanks for reading. Speak soon. Best, Marcus

📕 AI-Propaganda: Ragebait, Slop, and Synthetic Disinformation
📗 Participatory Propaganda: Actively involve Fans and Allies
📘 Data delivered by the TikTok API were false
📙 What else

Subscribe now

📕 AI-Propaganda: Ragebait, Slop, and Synthetic Disinformation

We are living in a time where actual photos taken in real life by real people on real cameras look and feel like having been AI generated. Baudrillard would have loved it. It may come to you as a cutesy autumn vibe or freaky rescue fantasy or as new propaganda form. The thing is: people care less and less if photos represent anything that has actually happened IRL.

"I don’t know where this photo came from and honestly, it doesn’t matter." wrote Amy Kremer, a Republican National Committee member representing Georgia, on X after an image’s synthetic provenance was revealed (NPR).

“This is the ‘fuck it’ era of AI slop and of political messaging, where AI-generated images are used to convey whatever partisan message suits the moment regardless of truth. For all of the fearmongering about the potential of deepfake-driven disinformation, we have seen time and time again that whatever message is being delivered doesn’t need to be supported with real—or even realistic—photos or videos to convey it” writes Jason Koebler for 404 Media.

“I thought this would be a bigger event in the history of images than the invention of perspective in the 15th Century. I still think that's true” tweets  Trevor Paglan. He has written and worked on computer vision and AI for ages.

And i would love to hear his thoughts on AI Slop –  low-quality, poorly generated content using generative artificial intelligence tools, often without proper oversight or refinement being used by e.g. Donald Trump (AP) – the Slop Candidate (Atlantic) – , Elon Musk (Independent), political campaigners in India (DW), politicians in Indonesia (CfR) and the German right-wing party AfD (Frankfurter Rundschau).

I am currently collecting data and use cases for a paper on multimodal AI-generated propaganda on TikTok and other as a part of my ongoing investigation of postdigital propaganda (#130).

📗 Participatory Propaganda: actively involve fans and allies

A couple of weeks ago i wrote an analysis of the European election campaigns of selected parties on TikTok for Friedrich-Ebert-Stiftung. One of the core results was that the politicians' sometimes great commitment did not always pay off in terms of high reach. The AfD, which had outsourced the distribution of content to fans and followers, achieved significantly higher interaction rates with fewer videos. 

Strategists of the party outsourced the dissemination of content and coordinated it in a public telegram group called “TikTok Guerilla” (Correctiv). With that they basically mimicked Russian propaganda strategies applied after the invasion of Ukraine (Vice /Media Matters). Instead of paying influencers they gave away iphones in a raffle. 

From now on, the goal of democratic parties should be to actively involve fans and allies in the dissemination of convincing communication. The concept of participatory propaganda as the deliberate and systematic attempt to shape perceptions and direct behaviour of a target audience while seeking to co-opt its members to actively engage in the spread of persuasive communications (Wanless and Berk 2021) should be studied and applied. 

The winter semester has just started. In the coming weeks i teach two courses at TH Köln – University of Applied Sciences on “Political communication on TikTok” and “TikTok Tactics – German Federal Election 2025” where we will further investigate successful strategies for political campaigning. And i hope to share some insights on “Memes as Modern Soft-Propaganda-Tools” at Europe’s first Political Tech Summit that takes place on January 25 in Berlin. 

📘 Data delivered by the TikTok API were false

Last week i was part of a data sprint with great colleagues from UvA Amsterdam. We set up TikTok persona accounts on cleared smartphones, trained them on different topics and later scraped their feeds using the tool Zeeschuimer on a web browser. 

While i have been hesitant using TikTok’ research API due to the documented shortcomings (The Research API falls woefully short) this field report has confirmed my apprehensions: Lessons from TikTok's API Issues During the 2024 European Elections. Quote: Basically, viewing and sharing metrics of data delivered by the TikTok API were false for several months and that just before the 2024 European Elections.

Once again i want to point to the TikTok Tools List and ask for help. If you know of another tool to investigate TikTok, please feel free to add it to the list. Thanks to the colleagues who helped built the list.

📙 What else

🍏 Reading latent values and priorities in TikTok's community guidelines for children International Journal of Cultural Studies

🍏 I Took A TikTok Class. What Happened Next Blew My Mind
Defector

🍏 TikTok wants to turn millions of Americans into paid shopping influencers
Rest of world

🍏 The Social-Media Influencers Reshaping How Young Americans Get Their Political News WSJ

🍏 ByteDance lays off hundreds of TikTok employees in shift to AI content moderation Techcrunch

🍏 Right-wing populist communication of the party AfD on TikTok: To what extend does the AfD use TikTok as part of its communication to win over young voters? The Dynamics of Digital Influence

And now let me point you to video artist Jordan Stone (TikTok) who is doing wholesome stuff that feels like some sort of polished, personal version of post-corecore. Here is an interview. 

I will be in Sheffield in a couple of days at the Association of Internet Researchers Conference 2024 presenting Tom Divon’s and my research on thirst trap propaganda on a fantastic panel called “Ambient Amplification: Attention Hijacking and Social Media Propaganda”, Friday, Nov 1, 3:30pm, SU Gallery Room 2. If you are around please come by and say “Hi”. Ciao ciao

Subscribe now

⚖️ To ban or not to ban

Is TikTok really going to be banned in the US? Nobody’s sure (Washington Post). TikTok argued against its U.S. ban in court, September 16 (NPR). TikTok says it's not spreading Chinese propaganda. The U.S. says there's a real risk (NBC). But is the hypothetical risk of Chinese spying through TikTok enough of a national security threat that it justifies a federal law that may infringe on Americans’ constitutional free-speech rights (CNN)? The Justice Department has asked that the three-judge panel issue a ruling by December. Legal experts expect the case to go to the supreme court, with the losing side in the federal court process expected to appeal against the ruling (Guardian).

Meanwhile the support for a U.S. TikTok ban continues to decline, and half of adults doubt it will happen (PEW Research Center).


⚡⚡Nazi Sound Propaganda

ISD researchers were able to find white supremacist content on TikTok (sample of 108 distinct videos uploaded by 75 unique users) within seconds of beginning their research. As they continued searching, the platform recommended more racist content to their accounts. With TikTok promoting search terms that help users evade content moderation (ISD).

Another investigation by the Data and Forensics team of Sky News was able to reveal 50,023 posts using sounds incorporating speeches from Nazis including Adolf Hitler and his chief propagandist Joseph Goebbels.The most popular sound made using a Hitler speech viewed by Sky News has been used in over 10,300 videos.

Most of the sounds using Nazi-related audio do not have explicit titles. However, when clicked on they instantly direct users to large numbers of posts using the same soundtrack. In this way, users link extreme content easily while minimising their exposure to moderation.

This practice is applied by other extremist actors on the platform too (#133). Learn more about The Sound of Disinformation in our paper: Bösch, M., & Divon, T. (2024). The sound of disinformation: TikTok, computational propaganda, and the invasion of Ukraine. New Media & Society, 26(9), 5081-5106.

🤩 Terminally Online

The Kamala HQ is doing a good job on TikTok (#132) thanks to ‘feral 25-year-olds’  some working their first jobs - given unfettered freedom to chase whatever they think will go viral (Washington Post).

The team has been experimenting with "content sludge" type ads that show two unrelated videos at the same time. Typically these videos are made for people with attention spans fried by TikTok. Or as 404 puts it: Kamala Harris Campaign Experiments With Ads for an Audience With “Brain Rot”.

Expect to see more of this. It basically just makes sense. Each medium comes with a fluid set of specifics and techniques that need fearless experimentation and evaluation. That is basically it. And then NATO's social team is reacting to @charli_xcx ending brat summer. What can you do?

🪆Russian Propaganda – 2024 Edition

There have been a bunch of leaks, investigations (e.g. DOJ), articles and studies on Russian propaganda and information warfare in the Post-Prigozhin era (Internet Research Agency). To quote Ellul: Propaganda is total. And that encompasses target audiences, platforms, techniques, narratives and all means involved. Why not use the comments sections or memes or or or.

Obviously there has been “a shift in tactics of Russian information operations, suggesting that instead of using bot & troll accounts with limited following, they’ve started supporting social media superspreader accounts with huge reach.” (Pekka Kallioniemi).

Besides these Russian state-affiliated accounts have accelerated their use of TikTok since the start of 2024 (Brookings). And this might be a good time to check again a joint investigation by DFRLab and BBC Verify that revealed a large network of fake TikTok accounts that spanned thousands of videos and hundreds of thousands of subscribers that garnered millions of views (EU Disinfo Lab).

The target audiences of these campaigns are inside and outside Russia, strategies may vary but it is worthwhile to examine different use cases like this investigation by Josef Šlerka and team: How Russian propaganda seeps into Czech TikTok.

And here is a report on Russian TikTok disinformation in Ukraine (Bloomberg) where “Russia is dominating us on TikTok due to the scale” of its operation, Kovalenko, who leads a department under the National Security and Defense Council, said in an interview in Kyiv. “The Russians have begun working systematically on TikTok and are utilizing this platform successfully.”

To find suited answers to this threat it is foremost important to understand specifics and techniques like applying sound and multimodal memes and outsourcing the dissemination through participatory propaganda.

What else?

🎾 Where are all the Black girls on TikTok? (Sage)
🎾 The New Public Intellectuals Are All on TikTok (Cosmpolitan)
🎾 TikTok Live is a strange and dystopian place (Taylor Lorenz)
🎾 Hat TikTok zum Erfolg der Afd beigetragen (Christian Pieter Hoffmann) {in german}
🎾 Medo Halimy is dead (AP)
🎾 Countries with the largest TikTok audience as of July 2024 (Statista)

Thanks for reading. Anyone in Sheffield or Manchester end of October? I am there for AoIR 2024. Just meet me in the alley by the railway station. Ciao

You are reading Understanding TikTok. My name is Marcus. I was in the midst of listening to an already autumn suited Charli XCX/Fleetwood Mac Mashup (Dreams) when P.E. Moskowitz’s text Vibeocracy hit me hard with these few succinct words: “And then, of course, it turned out that vibes did not mean much.”

Today we talk all about 🥷 terror 🏴 on the platform.

Subscribe now

TikTok alone will not turn you into a terrorist

The last couple of days in Germany have been all about a fatal knife attack in the west German city of Solingen (Islamic State group claims responsibility) that has placed immigration and Islamist terror at the top of the political agenda (FT) days ahead of key elections (Deutsche Welle). It took traditional media a couple of hours to headline: Tiktok terrorists and Islamism influencers: Almost all Islamist attackers have radicalized themselves online (NZZ). Where else, one might add.

TikTok terrorists (NZZ). A new compound word, made well known by political scientist and “terror expert” Peter R. Neumann (King’s College London). As a former journalist i understand the appeal of alliterations, yet i find this one especially imprecise.

It is undebatable that terror organizations and extremists try to persuade on any given platform. But the public discussion around TikTok especially in the context of terror seems rather uninformed while under-researched and pretty overheated to me.

Here is an example: A statement by Neumann in an interview with NZZ goes like this “if you click three times on a video that has to do with Palestine, for example, then you will almost only see corresponding content”. But current research by Anna Semenova, Martin Degeling, and Greta Hess shows that: “TikTok’s recommender system introduces content from diverse interests periodically to keep users engaged and help them discover new topics” (Auditing TikTok). 

And of course. Not everyone and not every criminal is a terrorist. Neumann himself points out in an interview: “by definition, a paranoid-schizophrenic violent criminal cannot be a terrorist”. And TikTok alone can not be made responsible for so-called “speed radicalisation” (Spiegel) whatever that might even mean, because radicalisation is a complex and timely process that can not be traced back to a singular application. Or as CNN reports: In light of an alleged terror plot on a Taylor Swift concert in Vienna, there is renewed concern about online extremism among youth. The 19-year-old alleged mastermind of the attack was radicalized online, Austrian authorities say, although it is unclear how. It is unclear. How.

Let’s have a more informed debate

Well. TikTok, right?! In order to lead a better informed discussion, here are some first observations concerning the who (bad actors), the how (strategies) and the whom (audience) in the context of islamist extremism (Verfassungsschutz) content on TikTok.

01) Audience: Terror suspects aged between 13 and 19
Terror organizations and extremists have diverse target audiences which they try to reach on a variety of suited platforms (e.g. Islamic State Content on Pinterest). In the center of attention has been the aspect of successfully targeting and recruiting teenagers. In an academic study of 27 ISIS-linked attacks or disrupted plots Neumann has revealed that of the 58 suspects, 38 were aged between 13 and 19 (CNN). In order to reach this specific target audience it of course makes sense to use the tools at hand. Close to a 100% of young adults in Germany use social media, 91,5% of them intensively, which means they use at least one platform multiple times a day (How young adults use TikTok). These numbers will most likely be very similar in many other countries.

👉What to do:
More education of and communication among pupils, students, teachers, parents and decision makers alike. Encourage critical media consumption: How can you spot a propaganda clip? Where is the difference between freedom of speech and hate speech? How to cross-check information on platforms?   

02) Bad Actors: From rock star preachers to execution footage
There is a variety of different bad actors in different parts of the world that are utilizing TikTok and other with different goals.

First reports about the Islamic State being active on TikTok for example date back to 2019 as WSJ reported: Islamic State Turns to Teen-Friendly TikTok, Adorning Posts With Pink Hearts. Research from The Institute for Strategic Dialogue found 11 Islamic State videos being shared openly on TikTok, including footage from executions in 2021: Hatescape: An In-Depth Analysis of Extremism and Hate Speech on TikTok. An investigation in 2023 on CaliphateTok: TikTok continues to host Islamic State propaganda showed an active Islamic State support network of 20 accounts with more than a million collective views.

There is wide range of actors and different approaches that can be identified. From Pro-Islamic State alternative news outlets that are operating multi-platform, multi-lingual disinformation operations under the guise of “media” (ISD) in Iraq to so called ‘rock star’ preachers influencing young people online (CNN) like Abul Baraa, a German-speaking Salafi preacher, described as “a gateway drug for more radical actors” or Ibrahim El Azzazi (Welt) or Raheem Boateng (Merkur).

👉 What to do:
More research here is needed to identify actors and strategies and to unwrap networks. Security authorities need a better understanding of platform specifics to better track and monitor content.

03) Strategies and Techniques: Coded Emojis, Sound Memes, and AI News
”Jihadi terrorists have historically adapted and leveraged new and emerging technologies for recruitment and operational ends”  (The Soufan Center). Bad actors embrace all variations of contemporary communication. They apply the full spectrum of postdigital-propaganda (#130).

This includes different tactics for different platforms: For Facebook, that includes replacing terrorist language with emojis. For Twitter, that involves toning down the content in English compared with what's posted in Arabic. For Telegram, it means copying directly from official ISIS material. It's an evolving cat-and-mouse battle with tech companies and national security agencies (Politico). Coded Emojis as dog whistles are common amongst rightwing extremists but islamist terrorists can do it too. Just watch out for the ninja and black flag emojis. Pretty obvious. I know.

The Islamic State supporters exploit TikTok's 'sounds' feature (Sky News): A TikTok user appears in a photograph holding two children. In another, he is in a gym, training with a worn-out punching bag. Visually, the TikTok post appears innocuous, but the audio is not. It consists of a speech given in Arabic, laid over an excerpt of vocal music known as a nasheed. This one was composed in support of Islamic State. Just a touch or click away is much more graphic content. Sound propaganda is effective and not that well known among fact checkers. Will get back to that in the next newsletter.

And - of course - there are various variations including AI. In 2024, ISIS supporters launched an AI-generated program called News Harvest to disseminate propaganda videos (The Soufan Center). The Global Network on Extremism & Technology reports about AI-Powered Jihadist News Broadcasts: A New Trend In Pro-IS Propaganda Production?

👉 What to do:
More research here is needed to learn more about strategies, campaigns, their target audience, reach and success. A knowledge transfer is needed to better inform fact-checkers, security authorities and politicians.

Islam ≠ Terrorism. I repeat. Islam ≠ Terrorism

These are some preliminary observations. I hope they help to lead a better informed, rational debate. And probably kick-off some research projects. After all TikTok is at the same time used for spreading islamist extremist and terrorist propaganda and for leading a critical discourse to respond to the binomial “Islam = terrorism” in order to challenge this misconception (see Social Media & Otherness).

What else?
Why Creators Have Stopped Editing Their Content (Rolling Stone)
Unpaid political content is thriving on the platform – and it is meme-infused (NYT)
Inside the Gen-Z operation powering Harris’ online remix (CNN)
We don’t have time for stories anymore. A vibe is enough (via Jules Terpak)
An influencer is running for Senate. Is she just the first of many? (Vox)
TikTok Creators and the Use of the Platformed Body in Times of War (SM&S)

Thanks for reading. Speak soon. Ciao