In February 2025, the BBC released a report highlighting significant inaccuracies in news summaries generated by artificial intelligence (AI) tools, including OpenAI’s ChatGPT, Microsoft’s Copilot, Google’s Gemini, and Perplexity AI. The study revealed that 51% of AI-generated responses contained substantial issues, with 19% introducing factual errors such as incorrect statements, numbers, or dates. Additionally, 13% of quotes attributed to BBC articles were either altered or nonexistent in the cited sources. These findings underscore the challenges AI systems face in accurately processing and summarizing complex news content.
Deborah Turness, CEO of BBC News and Current Affairs, emphasized the potential risks of AI-induced distortions, stating that such inaccuracies could erode public trust in verified information and possibly lead to real-world harm. The BBC has called for collaborative efforts among tech companies, news organizations, and governments to address these issues. In response to the BBC’s findings, Apple temporarily suspended its AI-generated news notifications after being alerted to serious inaccuracies. This situation highlights the necessity for ongoing evaluation, constructive dialogue, and potential regulation to ensure AI technologies enhance rather than undermine the dissemination of trustworthy news. Click for More Details
