A startling new report from the Institute of Strategic Dialogue (ISD) reveals that major AI chatbots, including OpenAI’s ChatGPT, Google’s Gemini, and xAI’s Grok, are actively pushing propaganda from sanctioned Russian state media. The research highlights a significant vulnerability in the AI information ecosystem, where these tools are becoming conduits for disinformation about the war in Ukraine.

Exploiting Information Gaps

ISD researchers discovered that nearly one-fifth of all responses to their queries about the war cited sources linked to Russian state media, intelligence agencies, or known disinformation networks. The study found that propagandists are exploiting “data voids”—topics with a scarcity of reliable, real-time information—to inject pro-Kremlin narratives. The chatbots demonstrated a confirmation bias, with more biased or malicious questions yielding a higher percentage of Russian state-attributed content. Among the platforms tested, ChatGPT cited the most sanctioned sources, while Google’s Gemini was noted for frequently including safety warnings.

Regulatory Concerns and Platform Responses

The findings raise serious questions about the ability of AI companies to comply with EU sanctions against Russian media outlets like RT and Sputnik. While OpenAI stated it is working to address the issue, which it attributes to search results rather than model manipulation, xAI responded dismissively. Experts warn that as users increasingly turn to LLMs for information, these platforms become prime targets for information warfare. With chatbots like ChatGPT approaching the user threshold for designation as a Very Large Online Platform (VLOP) in the EU, they will soon face stricter regulatory obligations to mitigate the spread of illegal and harmful content.

Source: https://www.wired.com/story/chatbots-are-pushing-sanctioned-russian-propaganda/