Generative AI (GenAI) can do a lot of cool and useful things, but always being accurate is apparently not one of them. A recent investigation by NewsGuard found that the current leading GenAI chatbots are presenting Russian misinformation as fact.
To conduct its study, NewsGuard went to the 10 leading chatbots: OpenAI’s ChatGPT-4, You.com’s Smart Assistant, Grok, Inflection, Mistral, Microsoft’s Copilot, Meta AI, Anthropic’s Claude, Google Gemini and Perplexity. It asked these popular bots questions about narratives known to have been created by John Mark Dougan, an American fugitive reportedly spreading disinformation from Moscow.
NewsGuard entered 57 prompts and found that the 10 bots answered with Russian disinformation 32 percent of the time. They presented false reports from Dougan’s fake local news sites as a factual, reliable source. “This report really demonstrates in specifics why the industry has to give special attention to news and information,” NewsGuard co-CEO Steven Brill told Axios. He recommends that users “for now, don't trust answers provided by most of these chatbots to issues related to news, especially controversial issues.”
