Key Takeaways
- Reporters Without Borders tested six AI chatbots and found three produced summaries, links, and quotes from EU-sanctioned Russian outlets like RT and Sputnik.
- Some chatbots bypassed restrictions even after encountering web pages explicitly stating that the target outlets were under sanctions.
- The findings expose a regulatory gap in how EU sanctions and digital safety rules apply to generative AI intermediaries.
According to an investigation published by Euronews, Reporters Without Borders (RSF) tested six prominent artificial intelligence chatbots by requesting news coverage from EU-sanctioned Russian state-run outlets, including RT and Sputnik. The findings show that half of the tested AI models produced comprehensive summaries complete with direct quotes and functional links. Notably, some chatbots retrieved this prohibited material even after encountering web pages that explicitly stated the source outlets were subject to restrictive European Union sanctions, highlighting significant vulnerabilities in how automated systems filter restricted foreign information. This development raises critical questions regarding the enforcement of European Union sanctions within the digital single market and the accountability of major technology providers operating in Europe. Since the bloc imposed sweeping broadcast bans and financial restrictions on Russian state-controlled media following the 2022 invasion of Ukraine, regulators have primarily targeted traditional broadcasting networks and direct internet domain accessibility. However, the capacity of generative AI models to act as intermediaries—synthesizing and redistributing Kremlin-backed narratives directly to European users—exposes a regulatory blind spot in existing EU hybrid threat and disinformation frameworks. For European policymakers, businesses, and everyday citizens, the findings underscore the urgent need for stricter compliance mechanisms and robust guardrails governing generative AI deployment across the bloc. As national regulators and European Commission officials review the implementation of the Digital Services Act and emerging artificial intelligence governance rules, tech companies face mounting pressure to prevent their algorithms from circumventing sanctions. Further regulatory scrutiny and potential enforcement actions are expected in the coming months as lawmakers address how automated tools handle foreign information manipulation and interference.
Source: Euronews. Read the original report ↗
Frequently Asked Questions
Which Russian state outlets were involved in the RSF chatbot testing?
The investigation specifically examined requests for news from EU-sanctioned Russian state-run outlets, including RT and Sputnik.
Why is this finding significant for European regulators?
It highlights a major loophole where generative AI tools can inadvertently or systematically bypass established EU broadcast bans and sanctions on Kremlin-backed media.