A study by the European Broadcasting Union and the BBC found that leading AI chatbots—including OpenAI’s ChatGPT, Google’s Gemini, Microsoft’s Copilot and Perplexity—produced at least one significant problem in 45% of more than 2,700 answers to news-related questions. The most common flaw was sourcing, affecting 31% of responses, followed by accuracy (20%) and lack of context (14%). Gemini posted the highest rate of significant issues at 76%, largely tied to sourcing, while all systems made basic factual errors, the report said. The companies did not immediately respond to requests for comment. EBU Deputy Director General Jean Philip De Tender and the BBC’s head of AI, Pete Archer, urged tech firms to prioritize reliability and publish results by language and market. Academics said the findings underscore the need for stronger media literacy among news consumers.
Related articles:
– Hallucination (artificial intelligence)
– Large language model
– Fact-checking
– EU Regulatory Framework for AI (AI Act)
– SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models
