🔍 New Investigation: How did Hugging Face's "Spaces," a platform designed for sharing open-source AI demos, become a hub for tools capable of generating non-consensual intimate imagery? 🧵
Marc Faddoul
@marcfaddoul.com
Director at @aiforensics.org - Algorithmic Audit and Platform Accountability non-profit Also with @freeourfeeds.com to expand ATproto infra and ecosystem!
Congratulations to our researcher Dr. Paul Bouchaud, who was awarded the Innovation Prize at the @viginum.bsky.social × Inria Scientific Prize. The prize recognized the paper “Beyond Guidelines: Assessing Meta’s Political Ad Moderation in the EU,” co-authored with VIGINUM.
We submitted 11 recommendations to Dutch lawmakers, including: -Support algorithmic pluralism -Treat ad library functionality as a standalone compliance obligation
🚨 NEW: Large-scale Telegram networks are spreading and monetizing non-consensual intimate images (NCII), including CSAM, in Italy and Spain. Our investigation reveals an organized ecosystem of abuse at scale — involving 25,000 users. 🧵
Last Monday, I found a dinner invitation from Emmanuel Macron in my junk mail. It was for the AI Summit anniversary, 24 hours later. Almost missed it! I jumped on a train and found myself at the Élysée the following evening. The room was filled with leading AI academics, startups, and big corps
🔒Safer Internet Day is a good moment to reflect on how our investigations are helping push the internet in a safer direction — and how we can continue being a catalyst for change in 2026. 🧵
Stay tuned! We’ve received many requests in recent days for updates on our Grok report. We’re currently conducting additional monitoring and analysis and will have more to share shortly. 🔍
As part of preparing for #ATmosphereConf in Vancouver March 2026, we're republishing the videos from Seattle March 2025, along with transcripts. Marc Faddoul, from @freeourfeeds.com, gives a talk at ATmosphere Conference 2025 in Seattle.
Marc Faddoul: Developing the AT infra ecosystem and governance - ATProto Community
ATmosphereConf Seattle 2025
atprotocommunity.leaflet.pub
⚠️ New Investigation: Grok is systematically “undressing” women and generating extremist content. Our latest research shows this isn’t just anecdotal, but a pattern at scale.
Let’s get to the point. In the EU, you have more control over what you see online. ℹ️link.europa.eu/qCfg8p
⚠️ AI chatbots are reshaping how we search. But EU rules haven’t caught up. Our new policy report digs into the risks of LLM-powered search and the gap in Europe’s regulatory approach. So how do we fix it? 👇 Full report + recommendations: tinyurl.com/2p2h77et
From 'Googling' to 'Asking ChatGPT': Governing AI Search
We provide an overview of the most relevant changes and shifts between traditional search engines and AI search functionalities. This report also offers a framework for situating AI search in the curr...
tinyurl.com
🎥 #TikTok, under influence? In this 2-part @artefr.bsky.social documentary, we explore a critical question: What is the role of TikTok as an instrument of geopolitical power? In the film, I focus on three things that make TikTok stand out today:
Europe is rapidly rolling out age-verification — in the name of “child safety.” 🚷 But is it actually working? Our first audit found AgeGO’s selfie feature streamed biometric data straight to Amazon. 🪪 Our new report examines AgeVerif: What we found raises serious concerns. 👇
The deployment of age verification methods on the internet is not without risk for users, whether they are adults or minors. These methods require the processing of highly sensitive data such as biometric markers or identity documents, which can then be opaquely exchanged between entities.
Paul Bouchaud@paulbouchaud.com · 11mo agoLast week, 5 major porn sites rolled out “age verification” in France to comply with regulators. 3 of them — xvideos, xnxx, tnaflix — use AgeGO. AgeGO: • Collects the exact video URL before consent • Sends webcam streams directly to Amazon • Forces users to give an email via an unskippable step
🚨 "AI Slop" investigation: Platforms failing to label synthetic content We found: - 1 in 4 TikTok search results contain AI imagery - Specialized "Agentic AI Accounts" mass-produce synthetic content - 80%+ of AI content is photorealistic but unlabeled - Instagram desktop users see zero AI labels
📢 In a joint civil society effort, we filed a formal DSA complaint against 𝕏 (formerly Twitter) for violating Article 26(3) — which bans targeting ads based on sensitive data like sexuality, politics, religion, health. 📄 Read the joint statement: aiforensics.org/uploads/Join...
aiforensics.org
Behind the scenes of a UN “AI for Good” conference sponsored by @microsoft.com: mentions of the complicity of BigTech in enabling the genocide in Gaza gets censored. Thank you Abeba for your bravery and stature. Definitely the best talk there was to be seen there
Dr Abeba Birhane@abeba.blacksky.app · last yr.genevasolutions.news/science-tech...
UN AI summit accused of censoring criticism of Israel and big tech over Gaza war - Geneva Solutions
A prominent AI scientist says she was pressured by the organisers of the UN’s flagship conference on AI to censor parts of her presentation that criticised Israel over its war in Gaza and the role of ...
genevasolutions.news
En 2024 et 2025, une agence pro-Orbán a mené une campagne numérique occulte en faveur du Rassemblement national. Des dizaines de contenus ont été diffusés sur X sur la base d’un ciblage politique (ce qui est interdit), selon des données de l’ONG @aiforensics.org partagées avec @mediapart.fr. 👇
Une agence pro-Orbán a mené une campagne numérique occulte pour le RN en France
Des contenus ont été diffusés sur le réseau social X en ciblant les internautes sur la base de critères politiques, ce qui est interdit par les réglementations françaises et européennes.
mediapart.fr
Wanna check if you've been targeted by ads based on sensitive data? Just drop your X portability archive to find out haveibeentargeted.online
In @aiforensics.org’s last study authored by @paulbouchaud.com we depict how advertisers use political and other sensitive user data to target ads on X aiforensics.org/work/sensiti...
aiforensics.org
Pixels | Le Monde@pixelsfr.bsky.social · last yr.X a permis à des annonceurs de cibler leurs publicités en fonction de critères politiques, une pratique interdite
🧵How TikTok failed Polish democracy during the 2025 Presidential Election Our new investigation exposes serious gaps in TikTok's election content moderation that disproportionately harmed the Polish diaspora : aiforensics.org/work/tiktok-...
Tiktok's Polish Elections Labels: Only Sometimes, And Only For Some
Our investigation reveals TikTok's systematic failures in applying election information labels during Poland's 2025 Presidential Election. The platform excluded 20+ million Polish diaspora users from ...
aiforensics.org
For Europe to regain sovereignty over its digital infrastructure, we also need sovereign funding for the ecosystem that defends its integrity! The Sunstorm Foundation was just launched to rally a new generation of European, tech-savvy, and entrepreneurially-minded philanthropists to fill this gap
lesechos.fr
The most entertaining and apropos thing the internet has to offer this week!
#TeslaTakedown@teslatakedown.com · last yr.Portrait artist @ianbrucestudio.bsky.social is doing a 6 day marathon livestream painting a portrait of Elon Musk. Check it out. www.youtube.com/live/uogmc-c... #TeslaTakedown #NoKings
We have two big announcements to share with you! free-our-feeds.ghost.io/free-our-fee...
Free Our Feeds - Update #2, May 2025
A lot has happened since our last update in February: loads of activity in the ATProto ecosystem and, of course looking at the bigger picture, the Trump administration’s policies have upended geopolit...
free-our-feeds.ghost.io
Ok my called shot: we should think about an ATproto App Store. 1. DMA affords alternative App Stores in the EU. 2. ATproto apps can provide other types of guarantees to users. Where can users learn about apps? Nowhere right now.
Yes! that was youchoose.ai an interesting proof of concept at the time
Feed Your Head, Choose Your Algorithm
Change your recommender system, get real GEMS from YouTube Content Creators
youchoose.ai
Ted Han★ 韓聖安@knowtheory.net · last yr.@marcfaddoul.com mentioning prior projects that were adversarial approaches to algorithmic choice like browser extensions to rerank YouTube feeds #ahoy25
"This is a highly adversarial space, we are investing a lot in our detection systems..." Meanwhile, our latest report at @aiforensics.org found Meta reviewed, approved, and displayed fake medicine advertisements over 292 million times to EU users since the DSA took effect.
