Google's Sundar Pichai advises users not to "blindly trust" everything AI tools say, acknowledging their susceptibility to errors and the importance of a rich information ecosystem beyond AI. This comes amidst criticism for inaccuracies in Google's own AI Overviews and broader concerns about generative AI relaying false information.

• Sundar Pichai, Chief Executive of Alphabet (Google's parent company), stated AI models are "prone to errors."
• Experts like Gina Neff, Professor of Responsible AI at Queen Mary University of London, urge Google to take more responsibility for AI accuracy, criticizing the company for asking users to fact-check its tools while releasing them.
• Google unveiled Gemini 3.0, its new consumer AI model, and its integration into search is part of the company's bid to remain competitive against services like ChatGPT.
• BBC research found that AI chatbots, including OpenAI's ChatGPT, Microsoft's Copilot, Google's Gemini, and Perplexity AI, inaccurately summarized news stories, with AI assistants still misrepresenting news 45% of the time despite improvements.