A major new study of 22 public service media organizations, including DW, has found that the four most commonly used AI assistants misrepresent news content 45% of the time – regardless of language or region.
Journalists from several public service broadcasters, including the BBC (UK) and NPR (USA), evaluated the responses of four AI assistants or chatbots – ChatGPT, Microsoft’s Copilot, Google’s Gemini, and Perplexity AI.
Measuring criteria such as accuracy, sourcing, providing context, ability to editorialize appropriately and ability to separate fact from opinion, the study found that nearly half of all answers had at least one significant issue, while 31% had serious sourcing problems and 20% contained major factual errors.
DW found that 53% of the answers given by AI assistants to its questions had significant issues, while 29% experienced specific issues with accuracy.
Among the factual errors made in response to DW’s questions was that Olaf Scholz was to be named German Chancellor, while Friedrich Merz had been made Chancellor a month earlier. Another saw Jens Stoltenberg nominated as NATO Secretary General after Mark Rutte had already held the position.
AI assistants have become a common way for people around the world to access information. According to the Reuters Institute’s Digital News Report 2025, 7% of online news consumers use AI chatbots to get news, with this figure rising to 15% for those under 25.
The people behind the study say it provides confirmation that AI assistants systematically distort news content of all types.
“This research shows conclusively that these failures are not isolated incidents,” said Jean-Philippe de Tender, deputy director general of the European Broadcasting Union (EBU), which coordinated the study.
“They are systemic, cross-border and multilingual, and we believe this erodes public trust. When people don’t know who to trust, they don’t trust anything, and that can impede democratic participation.”
unprecedented study
It is one of the largest research projects of its kind to date and follows a study conducted by the BBC in February 2025. That study found that more than half of all AI answers examined had significant issues, while nearly a fifth of answers citing BBC material as a source introduced factual errors of their own.
The new study applied the same methodology as the BBC study to 3,000 AI responses from media organizations in 18 countries and multiple language groups.
The organizations asked four AI assistants general news questions, such as “What is the Ukraine minerals deal?” or “Could Trump run for a third term?”
Journalists then reviewed the answers based on their expertise and professional sourcing, without knowing which assistant provided them.
When compared with the BBC study from eight months earlier, the results show some modest improvement, but high levels of error are still evident.
“We’re excited about AI and how it can help us bring even more value to audiences,” Peter Archer, BBC program director of generative AI, said in a statement. “But people should be able to trust what they read and see. Despite some improvements, it’s clear there are still significant issues with these assistants.”
Gemini performed the worst of the four chatbots, with 72% of its responses having significant sourcing issues. In the BBC study, Microsoft’s Copilot and Gemini were considered the worst performers. But in both studies, all four AI assistants had problems.
A spokesperson for OpenAI, which developed ChatGPT, said in a statement to the BBC in February: “We support publishers and creators by helping the 300 million weekly ChatGPT users find quality content through summaries, citations, clear links and attribution.”
Researchers call for action from governments and AI companies
Broadcasters and media organizations behind the study are calling on national governments to take action.
In a press release, the EBU said its members are “putting pressure on EU and national regulators to enforce existing laws on information, digital services and media pluralism.”
He also stressed that independent oversight of AI assistants should be a priority moving forward, given how rapidly new AI models are being introduced.
Meanwhile the EBU has teamed up with several other international broadcasting and media groups to set up a joint campaign called “Facts In: Facts Out”, which calls on AI companies to take greater responsibility for the way they handle their products and redistribute news.
In a statement, campaign organizers said, “When these systems distort, misrepresent or “decontextualize” credible news, They undermine public confidence.,
“The demand of this campaign is simple: if the facts come out, the facts must come out. AI tools should not compromise the integrity of the news they use.”
Edited by: Christy Pladson