Lightnews — Scholar-powered news

AI Forensics @aiforensics.org · Jul 31

Full report: aiforensics.org/work/gen-ai-...

This investigation systematically analyzes AI-generated content across TikTok and Instagram in three European countries. We found that 25% of TikTok's top search results contain synthetic AI imagery, ...

aiforensics.org

1 1 3

AI Forensics @aiforensics.org · Jul 31

Platforms acknowledge AI as DSA risk but aren't acting on it.

www.spiegel.de/netzwelt/ai-...

(S+) TikTok und Instagram: Wie KI-Schrott die Timelines auf Instagram und TikTok verstopft

Im Netz breiten sich massenhaft generierte KI-Inhalte rasend schnell aus. Das belegt eine neue Studie. Das Forscherteam warnt vor Gefahren für die Demokratie.

www.spiegel.de

1 2

AI Forensics @aiforensics.org · Jul 31

🚨 "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

1 2 9

AI Forensics @aiforensics.org · Jul 15

@cdt.org , Entropy, European Digital Rights (Edri), @freiheitsrechte.org @globalwitness.org @panoptykon.eupolicy.social.ap.brid.gy @bitsoffreedom.bsky.social @voxpublic.bsky.social

3

AI Forensics @aiforensics.org · Jul 15

📢 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

1 9 15

Reposted by AI Forensics

Paul Bouchaud @paulbouchaud.com · Jun 19

Wanna check if you've been targeted by ads based on sensitive data? Just drop your X portability archive to find out haveibeentargeted.online

Screenshot of the website haveibeentargeted.online displaying the result of the analysis. We observe that the user was served ads based on their inferred similarity with political accounts.

3 6

AI Forensics @aiforensics.org · Jun 18

Our investigation covered on @pixelsfr.bsky.social by @damienleloup.bsky.social !

Pixels | Le Monde @pixelsfr.bsky.social · Jun 18

X a permis à des annonceurs de cibler leurs publicités en fonction de critères politiques, une pratique interdite

La méthode, proscrite en Europe, sauf rares exceptions, par le règlement général sur la protection des données personnelles, reste utilisée, voire encouragée, sur le réseau social d’Elon Musk, révèle une étude de l’ONG AI Forensics.

www.lemonde.fr

4

Reposted by AI Forensics

Paul Bouchaud @paulbouchaud.com · Jun 18

How X enables major brands to target your most sensitive personal data - from health conditions to political beliefs - check out our new report! aiforensics.org/work/sensiti...

1 2 2

AI Forensics @aiforensics.org · Jun 18

We also came up with haveibeentargeted.online
where you can upload your X data and see whether advertisers have targeted you based on your sensitive characteristics 👇

AIF - Have I Been Targeted ?

haveibeentargeted.online

1 1

AI Forensics @aiforensics.org · Jun 18

💡 Silver lining: X's transparency exposed what other platforms may hide. We're calling on Instagram, TikTok, and others to match this level of disclosure.

2

AI Forensics @aiforensics.org · Jun 18

📊 The scale is massive: These ads generated hundreds of millions of impressions across the EU, creating a system where your most personal characteristics determine what information you see.

It's algorithmic discrimination at industrial scale.

1 2

AI Forensics @aiforensics.org · Jun 18

🌍 DISCRIMINATORY TARGETING: Saudi Arabia's Public Investment Fund excluded users based on:
• Ethnicity (#Blacks, #Niger)
• Sexual orientation (#gay, #lesbian)
• Faith (#Jewish, #Jews)

1

AI Forensics @aiforensics.org · Jun 18

🏛️ Even worse: EU institutions violated their own rules. The European Commission illegally targeted citizens using political data—confirmed by the European Data Protection Supervisor after a complaint by @noybeu.bsky.social

1

AI Forensics @aiforensics.org · Jun 18

💊 HEALTH TARGETING: Dell Technologies targeted users with specific medication names like "Sertraline," "Diazepam," and "Gabapentine".

McDonald's excluded users based on mental health terms like "suicidal" and "depressed."

Your medical data = their marketing advantage.

1 1

AI Forensics @aiforensics.org · Jun 18

🗳️ Political Targeting: from TotalEnergies excluding users interested in Green politicians and environmental activists to Netflix or Société Générale that excludes those interested in French far-right party.

1

AI Forensics @aiforensics.org · Jun 18

Compared to Meta, X is more transparent in the targeting parameters used for advertisings, and such disclosure, mandated by the DSA, reveals how the platform actively enables discriminatory advertisings.

1

AI Forensics @aiforensics.org · Jun 18

We analyzed X's Ads Repository and discovered that **X's allowed widespread targeting of users over "sensitive categories" protected under EU law.

1

AI Forensics @aiforensics.org · Jun 18

🧵How X enables major brands to target your most sensitive personal data—from health conditions to political beliefs—all in the name of profit.

aiforensics.org/work/sensiti...

aiforensics.org

1 6 6

Reposted by AI Forensics

Sleeping Giants FR @slpng-giants-fr.bsky.social · Jun 18

💥 Des pubs sur X ciblent sur l'orientation politique, syndicale, religieuse ou les préférences sexuelles.
C'est interdit en Europe.
X fournit même clé en main des catégories de ciblage potentiellement illégales que ses clients utilisent.
Bravo à @aiforensics.org pour l'excellent travail d'analyse !

X a permis à des annonceurs de cibler leurs publicités en fonction de critères politiques, une pratique interdite

La méthode, proscrite en Europe, sauf rares exceptions, par le règlement général sur la protection des données personnelles, reste utilisée, voire encouragée, sur le réseau social d’Elon Musk, révèle ...

www.lemonde.fr

2 76 120

Reposted by AI Forensics

Marc Faddoul @marcfaddoul.com · Jun 18

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...

2 7

AI Forensics @aiforensics.org · Jun 13

If we can't independently verify what platforms share through APIs, "transparency" becomes a mere illusion.

Real transparency means researchers can independently verify platform claims, not just accept whatever data companies choose to share.

2

AI Forensics @aiforensics.org · Jun 13

We built a public dashboard monitoring 10 high-profile videos that should be accessible but aren't.

Check it yourself: playground.tiktok-audit.com/api-na/

TikTok Research API - Availability Dashboard

playground.tiktok-audit.com

1 2

AI Forensics @aiforensics.org · Jun 13

This destroys research validity. Studies using "data donations" (where users share their TikTok history) can't analyze 1 in 8 videos participants actually saw.

Algorithm audits, content analysis, platform accountability research—all compromised by hidden data.

1 1

AI Forensics @aiforensics.org · Jun 13

This isn't just about numbers. When the TikTok CEO posted a response to the US TikTok ban with 30+ million views, researchers studying platform political communication couldn't access it.

How can we hold platforms accountable if they control what we can study?

1 1

AI Forensics @aiforensics.org · Jun 13

Result?

After checking which videos were actually public, we found 1 in 8 videos were missing for no valid reason.

TikTok's own company videos (CEO statements, official announcements), major accounts Taylor Swift, Brook Monk, and news outlets.

All in all, ~1% of creators completely excluded.

1 1