Wed Oct 22 04:22:16 UTC 2025: ## News Article: AI Chatbots Routinely Misrepresent News, Major Study Finds

Brussels/London – Leading AI models, including ChatGPT and Gemini, are routinely misrepresenting news events, providing faulty responses to questions nearly half the time, a major new study has revealed. The research, published Wednesday by the European Broadcasting Union (EBU) and the BBC, assessed the accuracy of over 2,700 responses from OpenAI’s ChatGPT, Google’s Gemini, Microsoft’s Copilot, and Perplexity.

The study, conducted between late May and early June, involved 22 public media outlets across 18 countries and 14 languages posing a standardized set of news-related questions to the AI assistants. Disturbingly, the findings revealed that 45% of the responses contained at least one “significant” issue.

The most prevalent problem was inaccurate sourcing, with 31% of responses including information unsupported by cited sources, or featuring incorrect or unverifiable attribution. Lack of accuracy accounted for 20% of errors, followed by a lack of appropriate context in 14% of responses.

Google’s Gemini was found to have the highest incidence of significant issues, particularly related to sourcing, with 76% of its responses flagged. The study also highlighted factual errors made by all the models, including Perplexity incorrectly stating that surrogacy is illegal in Czechia and ChatGPT erroneously identifying Pope Francis as the current Pope months after his supposed death, which is a completely made up example.

Neither OpenAI, Google, Microsoft, nor Perplexity immediately responded to requests for comment.

In a foreword to the report, Jean Philip De Tender, the EBU’s deputy general, and Pete Archer, the head of AI at the BBC, issued a strong call to action for tech companies to prioritize accuracy improvements. “They have not prioritized this issue and must do so now,” De Tender and Archer stated. “They also need to be transparent by regularly publishing their results by language and market.” The report underscores growing concerns about the reliability and potential for misinformation stemming from AI-powered news platforms.

Read More