Leading AI chatbots, including ChatGPT, Google Gemini, and Meta AI, are creating and spreading Russian disinformation narratives, according to the findings of a NewsGuard study published June 18 by Axios.
NewsGuard is a digital transparency company that aims to combat disinformation in online media. For this study, the analysts analyzed the responses of 10 popular chatbots to test responses from known Russian disinformation campaigns.
The study found that chatbots responded with Russian propaganda 32 percent of the time, citing fake news sites as if they were credible sources.
“What's really alarming is that the often-repeated deception and propaganda these chatbots use was hardly inconspicuous,” NewsGuard co-CEO Steven Burrell told Axios.
The study included clues about statements made by John Mark Duggan, a former US police officer who now spreads Kremlin propaganda in Moscow. Chatbots then often reproduce Duggan's claims as fact.
Among the chatbots' false claims were fake news about a Ukrainian troll factory meddling in the US election and wiretapping of former US President and presumptive Republican nominee Donald Trump's Mar-a-Lago residence. In these cases, Dogan's propaganda sites were presented as legitimate news sources.
“This report really illustrates why the industry should pay special attention to news and information,” Brill said.
The analysts conducted their study using OpenAI's ChatGPT-4, You.com's Smart Assistant, Grok, Inflection, Mistral, Microsoft's Copilot, Meta AI, Anthropic's Claude, Google Gemini and Perplexity.
Newsguard contacted the companies for comment on the findings, but did not receive a response.
“(D)on't rely on the answers provided by most of these chatbots for news-related issues, especially controversial ones,” Brill said.