Beware of AI Applications: Widespread News Distortion

New European Research Reveals Details

Beware of AI Applications: Widespread News Distortion

New research published by the European Broadcasting Union and the BBC on Wednesday shows that leading AI applications distort news content in nearly half of their responses. 

The international research examined 3,000 responses to news questions from AI assistants, software applications that use AI to understand natural language commands to complete tasks for the user.

The study evaluated AI assistants in up to 14 languages ​​for accuracy, sources, and the ability to distinguish between opinion and fact. These assistant applications included ChatGPT, Copilot, Gemini, and Perplexity. Overall, the research showed that 45% of the AI ​​responses studied contained at least one major problem, and 81% of them contained some form of problem. Reuters contacted the companies for comment on the findings.

OpenAI and Microsoft previously said they are working to solve the problem of hallucination, when an AI model generates incorrect or misleading information, often due to factors such as insufficient data. Perplexity states on its website that one of its "deep search" models has a 93.9% accuracy rate in terms of factual accuracy.

Source Errors

The study showed that a third of AI assistants' responses showed serious source errors, such as missing, misleading, or incorrect attribution. The study said that about 72% of responses from Gemini, Google's AI assistant, had significant source issues, compared to less than 25% for all other assistants. The study said that accuracy issues were found in 20% of responses from all AI assistants studied, including outdated information.

Examples cited in the study include Gemini incorrectly reporting changes to the law on disposable e-cigarettes, and ChatGPT incorrectly identifying Pope Francis as the current pope months after his death. The study involved 22 public service media organizations from 18 countries, including France, Germany, Spain, Ukraine, the UK, and the US.

The European Broadcasting Union (EBU) said that as AI assistants increasingly replace traditional news search engines, public trust could be undermined. Jean-Philippe de Tendre, EBU's media director, said in a statement: "When people don't know what to trust, they end up not trusting anything at all, and this can hinder democratic participation."

The Reuters Institute's Digital News Report 2025 indicated that approximately 7% of all online news consumers and 15% of those under 25 use AI assistants to get news. The new report urges AI companies to be accountable and improve how AI assistants respond to news inquiries.