As artificial intelligence becomes increasingly embedded in daily life, OpenAI CEO Sam Altman is sounding a note of caution. On a recent episode of the OpenAI Podcast, Altman urged users not to place blind faith in ChatGPT, reminding the public that the popular tool can still produce false or misleading information.
“It should be the tech you don’t trust that much,” he said. “AI hallucinates.” According to Altman, it’s precisely because of this tendency that users must learn to use the tool responsibly, without assuming every output is accurate.
Altman offered a personal reflection, admitting that he leaned too heavily on ChatGPT after becoming a new parent. From nap schedules to diaper rash remedies, he used the AI to help make parenting decisions, until he realized it wasn’t always reliable. “I had to remind myself,” he said, highlighting how even tech leaders can fall into the trap of over-dependence.
Beyond individual habits, Altman stressed the broader implications of mass reliance on AI systems. He called for clear societal frameworks to help regulate how tools like ChatGPT are used. “We need societal guardrails,” he said, urging users to keep human judgment and real-world context at the core of any AI interaction.
