A significant study has unveiled that over 50% of news summaries produced by AI chatbots, including OpenAI’s ChatGPT, Microsoft’s Copilot, and Google’s Gemini, were found to contain inaccuracies. Led by the BBC and spanning 22 public service media organizations, the European Broadcasting Union (EBU) study scrutinized thousands of responses, revealing issues in nearly half of them. Google Gemini was notably flagged for its 76% error rate, primarily due to sourcing inaccuracies. The study highlighted the risk of misinformation, with chatbots often displaying confident but unfounded assertions. To address these challenges, a toolkit has been developed to help developers refine chatbot accuracy, underscoring the importance of precision in digital news delivery for maintaining public trust.