Graphika's team is what makes our company such an incredible workplace. And in our recent Great Place To Work certification, we learned that the vast majority of our employees agree. Interested in learning more about working at Graphika? See our Careers page: graphika.com/careers

Someone is in Kitteh heaven.(Bliss) #CatsofBluesky

Fuzzy beanz #CatsofBluesky

Looking good for 14 #Catsofbluesky

Walks in nature...

Marta Havryshko, propagovaná a obhajovaná niektorými slovenskými a českými "akademikmi" a propagovaná niektorými médiami ako Marker, Pluska, Alarm, Hlavné Správy.. Obľubená aj v odborárskych kruhoch. Čítajme Vatnik Soup. 1/2

Slovensko potrebuje Hlídacího Psa. hlidacipes.org/divka-z-veci...

"Zodpovednosť, nie náboženské prenasledovanie." Krátky článok (názor) o Cirkvi zjednotenia/Moonovcoch (a AllatRA) od experta, ktorý má so sektou osobnú skúsenosť. Verím, že si to prečíta napríklad Ján Figeľ alebo Marek Krajčí. www.koreatimes.co.kr/opinion/2026...

🔄 Recap of DE-CONSPIRATOR Annual Consortium Meeting On 17–18 December 2025, the consortium gathered at Özyeğin University, Istanbul, and brought partners together to align priorities as the project enters its final year. 🔗Read the the full article on our website: deconspirator.eu/de-conspirat...

Missed our latest webinar? You can still access Ask An Expert: When PR Looks Like Press... Tracking China's Spoofed Media Domains. Watch to learn about Graphika’s latest investigation into a sprawling network of spoofed media domains linked to China-based actors. graphika.com/reports/when...

Our FOSDEM Hackathon begins! Couldn't make it? Joins us at 17:00 CET on YouTube for a Matrix Live, actually Live from Brussels, and discover what attendees created or struggled with! youtube.com/live/U_Ydrcr...

🔍 DE-CON Glossary #14: Kompromat Kompromat is a Soviet-origin tactic involving the collection and use of compromising information to discredit, blackmail, or manipulate targets—and today extends far beyond Russia. 📌Definition from the DE-CONSPIRATOR FIMI Glossary (D2.2).

Element co-founder and COO, @amandine.bsky.social, participates in @openforumeurope.org's panel debate on Europe’s Software Challenge, discussing how openness underpins digital sovereignty.

The Hackathon is making progress, people have formed teams! Integrations, Spec, Custom Emojis, Encrypted Search... there are plenty of ideas flying around!

Thanks to our sponsor Famedly, people at the hackathon can get refreshed. We even get to play with a fancy bar code scanner!

Global risk reports agree on a clear message: FIMI, disinformation and misinformation have become systemic threats to democracies, economic stability and social cohesion. Strong public-interest media are a critical defence. Read more 👇

Not only do we have drinks, Famedly is also graciously sponsoring our lunch. And what better meal than pizzas to get more hacker brain juice?

TUA Research exposes 'InfoDefense' network of ~30 TG channels in multiple languages, pushing Russian propaganda, attacking Western govts & smearing mainstream media under the guise of a "grassroots" movement. We identify the leader, Gazprom engineer Alexandr Chernishov

Ukrainian artist, Zoriana Petriv

Plenty of conversations have been going on during lunch. Some serious, some less, but it's a very light hearted hackathon. Community first!

We're starting to see the first results of the hackathon! Florian here is building a Rust bot that talks to Christian's receipt printer, and makes it print what they send to a room. Prototypes are taking form, keep up the good work everyone!

Rýchla rota na scéne.

QuadRadical ismaking steady progress on Nexus, MatMaul is hooking encrypted search into Element X... the prototypes are polishing up!

A pair of cats diligently guards the local technology on #CaturdayEve.

Time is up for our hackathon! Our mighty teams have successfully hacked their way towards the Matrix Caps! We'll start live streaming the results in a few minutes! www.youtube.com/live/U_Ydrcr...

After Russia’s strikes on Ukraine’s energy sector and blackouts during severe frosts, Russian propaganda faced the problem of how to present these crimes to the world. In this article, we showed how they manage to present war crimes as everyday problems. #Russia #war texty.org.ua/articles/116...

It's time for the mission report, let's start the YouTube LiveStream! www.youtube.com/live/U_Ydrcr...

‼️‼️‼️‼️‼️‼️‼️‼️‼️‼️ 🚨SECURITY RISK🚨 🚀REPOST🚀 Planet🌏Anthro @planet_anthro on Xitter has been hijacked by a scammer. 🤬 I’m locked out of my account there. Can’t retrieve it. 😭 It is DMing extortion attempts!!! 😡 DO NOT CLICK ANY LINKS IN DMS‼️ DO NOT ENGAGE‼️

Follow TUA research #bluesky

Check out the new dispatch from the MAAZ team: movementagainstantizionism.substack.com/p/dispatch-1...

It is estimated that Yang and Jusufoski could be earning thousands if not tens of thousands of dollars a month, while the amount they pay for AI tools that create their videos for them are less than $50 a month.

Until then, those like Jusufoski and Yang will continue to be able to AI slop farm, using minimal seed money to create minimum effort content and relying on an unscrupulous algorithm to maximise their return on investment.

The question of how to find authentic content while algorithms decide to instead boost mass produced AI slop remains, and with SM platforms unlikely to act in anyone’s interest but their own, it may come down to whether governments have the political willpower to step in.

And social media companies realise the dangers of this kind of content on their platforms, with YT CEO Neal Mohan promising to use Ai to crack down on AI slop, although IG head Adam Mosseri seems to think users will naturally drift away from AI slop without platform intervention.

These “just for fun” channels are being promoted by Kremlin propagandists and info ops because the constant bombardment of cheap and effortless pro-Russian content is effective; not at making people die hard Putin supporters, but by subtly discrediting Ukraine, NATO and the EU.

These aren’t fringe actors, but algorithm boosted high impact channels whose content has been picked up by Vladimir Solovyov, Simeon Boikov, Anatolij Sharij and Alina Lipp as well as familiar faces such as InfoDef, Kat Lubek, Chay Bowes and Gerry Nolan of The Islander.

Jusufoski’s channels follow the same blueprint as Yang’s. AI slop shorts, posted daily, amplifying pro-Russian narratives such as Zelenskyy begging for funds, trivializing serious issues, distracting from Russian war crimes and softening the image of dictators like Putin.

The three other accounts - Vrhunske Priče, Hatim's Shorts and AI Comedy - are all run by Hatim Jusufoski. Hatim is North Macedonian, but despite being a fan of Putin, migrated to St. Gallen canton, Switzerland to get married and have a child in a liberal democracy instead.

Yang's accounts post truly lowest common denominator AI slop, minimal quality but designed for maximum, broad appeal, sacrificing informative value for high engagement and reach. The messaging is often little more than celebrating certain world leaders at the expense of others.

These basic, 9 second long videos rack up tens of millions of views each, as YouTube's algorithm rewards users that upload as frequently as possible and it takes mere minutes to create each video using platforms such as MemeGen AI and Higgsfield and models like Kling and Hailou.

The first of these channels is Make AI Great Again, which has a combined 1.6 million followers over YT, IG and TT, and is run from Adelaide, Australia by Chinese national Hu Yang.

While they may be difficult to take seriously by educated audiences, short videos depicting Zelenskyy dressed as a pig being eaten by Putin, fighting with a slingshot or stealing or begging for money from Western leaders reach audiences ever declining traditional media can't.

At the suggestion of and in collaboration with an anonymous Swedish contributor, TUA investigated 4 pro-Russian AI slop social media channels on YouTube, Instagram, Facebook and TikTok. Between them they have >3.5 million followers. On YT alone they have >2 billion views. 🧵

The majority of our research will also still be available via our LinkTree, which you can find at linktr.ee/TUA_Research

The choices we make as both consumers of and content creators on social media have never been more important, which is why we - after our next big thread - are switching to Bluesky. We will post here first, and only on X after. Follow us here to get our ID threads first.

Great decision! Worth a follow!