A European Broadcasting Union report reveals significant reliability issues with widely used AI assistants, including OpenAI’s ChatGPT, Microsoft’s Copilot, Google’s Gemini, and Perplexity. The study, which involved public media outlets across 18 European countries, found that nearly 45% of AI-generated answers contained at least one major error. The most concerning issue was outdated information, with AI frequently confusing current facts with inaccuracies or parodies. Gemini had the highest inaccuracy rate, with 76% of its responses considered flawed, primarily due to poor sourcing. Notably, even basic questions about public figures, such as the Pope, yielded misleading answers; for example, some AI claimed Pope Francis was still alive despite his replacement by Leo XIV. This highlights that AI assistants remain unreliable sources for current news, prompting experts from the EBU and BBC to caution against using them for accurate information.
Source link
