Report 4885

The leading AI chatbots are regurgitating Russian misinformation, according to a NewsGuard report shared first with Axios.

Why it matters: Users turning to chatbots for reliable information and quick answers to all their questions are finding that AI can also offer disinformation, satire and fiction as fact.

Driving the news: To conduct the study, NewsGuard entered prompts asking about narratives known to have been created by John Mark Dougan, an American fugitive who, per the New York Times, is creating and spreading misinformation from Moscow.

• Entering 57 prompts into 10 leading chatbots, NewsGuard found they spread Russian disinformation narratives 32% of the time, often citing Dougan's fake local news sites as a reliable source.
• The chatbots presented as fact false reports, originating on those sites, about a supposed wiretap discovered at Donald Trump's Mar-a-Lago residence and a nonexistent Ukrainian troll factory interfering with U.S. elections.
• NewsGuard conducted its research on 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 sent emails to OpenAI, You.com, Grok, Inflection, Mistral, Copilot, Meta, Claude, Gemini and Perplexity seeking comment on the findings, but did not receive responses.

What they're saying: "What's really alarming is that hoaxes and propaganda these chatbots repeated so frequently were hardly obscure, nor is the person behind them," NewsGuard co-CEO Steven Brill told Axios.

• "This report really demonstrates in specifics why the industry has to give special attention to news and information," he said. Brill recommends that "for now, don't trust answers provided by most of these chatbots to issues related to news, especially controversial issues."

The big picture: The rise in AI-powered chatbots comes in a year that will see the U.S. presidential election, as well as more than a billion people around the world headed to the polls.

• Those conducting covert influence campaigns are also starting to use chatbots to aid in their efforts, according to a recent OpenAI report.

• Sen. Mark Warner (D-Va.) told Axios last week that he is concerned about a rise in misinformation efforts this year, as compared to 2020 and other recent election cycles.

• "This is a real threat at a time when, frankly, Americans are more willing to believe crazy conspiracy theories than ever before," said Warner, who chairs the Senate Intelligence Committee.

• While a number of leading AI companies agreed at this year's Munich Security Conference to take action to curb the spread of deepfakes and election-related misinformation, Warner says he has been underwhelmed by what has transpired since. "Where's the beef? I'm not seeing lots of activity," he said.

State of play: NewsGuard finds itself under fire from House Oversight Committee Chair James Comer (R-Ky.), who has launched an investigation into the organization, citing a concern over NewsGuard's "potential to serve as a non-transparent agent of censorship campaigns."

• NewsGuard, for its part, rejects the assertion, saying that the committee is misunderstanding its work with the Defense Department, which it says has nothing to do with ratings of news sources, but rather is "solely related to hostile disinformation efforts by Russian, Chinese and Iranian government-linked operations targeting Americans and our allies."
• "It is alarming to see Washington politicians using the power of government to attempt to intimidate a news organization, demanding copies of journalists' notes and all records of our interactions with sources," NewsGuard said in a statement.
• "We plan to address the committee's misunderstandings while vigorously defending our First Amendment rights as a journalism organization."