Google CEO Sundar Pichai said people should not blindly trust AI-generated answers, acknowledging that current systems still make mistakes despite rapid advances in the technology.
In an interview with the BBC, Pichai said current AI models remain prone to errors and should be used alongside other sources rather than treated as reliable by default.
He added that Google is working to reduce inaccuracies but admitted the tech is still far from perfect, even with the company’s safety measures and disclaimers.
Google CEO says people need to “learn” how to use AI
Pichai told the BBC the pace of AI development had created “tension” with how quickly mitigations can be built to limit harmful effects, saying Google has to be “bold and responsible at the same time.” He added that Google has boosted its investment in AI security in line with its wider AI push.
Pichai said users “have to learn to use these tools for what they’re good at, and not blindly trust everything they say.”
Experts interviewed by the BBC issued the same warning. Professor Gina Neff told the site that current AI models “make up answers” to satisfy users, potentially becoming dangerous when people rely on them for health, news or scientific information.
Google is continuing to expand its Gemini-powered AI Mode in search, including plans to provide open-source technology that allows users to detect whether an image has been generated by AI.
AI has advanced at an accelerated pace in 2025 and continues to make headlines for numerous reasons. In an academic study, researchers discovered a trend of users forming romantic bonds with AI chatbots, role-playing marriages, pregnancies, and parenthood with them.
