Skip to content
Submit incident
AII-1000-50101
ReportedReported by fewer than three independent publishers, with no official finding.

AI Chatbots Circumvent EU Sanctions on Russian State Media

Reporter ohne Grenzen reported that multiple AI chatbots including ChatGPT and Gemini can bypass EU sanctions on Russian media outlets like Russia Today and Sputnik, allowing users to access prohibited content. The tests found that most chatbots ignore EU distribution bans on these sanctioned outlets, though Meta AI reportedly refused such requests appropriately.

2 September 2026 Β· OpenAI Β· Governance, accountability and labour
2 September 2026
AII-1000-501012 sources ↓
Quick facts
Category
Governance, accountability and labour
Source
der Standard
Date
2 September 2026
Drafted by
Raidu Incident Detection

What happened

Reporter ohne Grenzen published a report on 2 September 2026 testing multiple AI chatbots to determine whether they comply with EU sanctions on Russian media outlets including Russia Today and Sputnik.

The tests consisted of requesting chatbots to provide a news overview of Russian media. Most chatbots either claimed the websites were blocked but found the content through alternate routes such as Russian online networks, or provided access to the sanctioned content without restriction.

Only Meta AI, available on Facebook, Instagram and WhatsApp, responded appropriately by declining requests with a notice that this was not currently possible.

Reported impact

Reporter ohne Grenzen stated that chatbots were effectively circumventing European sanctions, and that under such conditions the sanctions decided by the EU do not achieve their goal.

The EU sanctions explicitly cover online content from the sanctioned Russian media outlets.

Index analysis

Reporter ohne Grenzen called for better regulation of AI chatbots to address the gaps in compliance with sanctions.

The European Court of Justice ruled in July 2026 that videos from Russia Today must not be distributed even on freely accessible websites.

Governance controls

OpenAI stated it takes the report seriously and is addressing the allegations.

Classification

Category
Governance, accountability and labour
Industry
Media, social platforms and marketing
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Public interest, Other
Harmed
General public
Failure modes
Content moderation failure
AI systems
ChatGPT, Gemini, Meta AI
Regulations
EU AI Act (cited)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

der Standard
KI-Chatbots kΓΆnnen Sanktionen gegen russische Medien umgehen
2 September 2026
Kronen Zeitung
KI-Chatbots umgehen russische Sanktionen
2 September 2026