HomeNewsTrends'People trust ChatGPT too much': OpenAI CEO Sam Altman warns AI still hallucinates

'People trust ChatGPT too much': OpenAI CEO Sam Altman warns AI still hallucinates

Sam Altman’s remarks came amid growing public reliance on artificial intelligence tools, especially large language models (LLMs) such as ChatGPT, for tasks ranging from professional research and customer service to parenting advice and creative writing.

June 26, 2025 / 17:56 IST
Story continues below Advertisement
OpenAI removes Sam Altman's ownership of its Startup Fund
Sam Altman himself admitted to using ChatGPT extensively for parenting guidance during his son’s early months.

OpenAI CEO Sam Altman, in a candid admission during the first episode of OpenAI’s newly launched podcast, said that people place a surprisingly “high degree of trust” in ChatGPT—despite the well-known fact that the AI frequently hallucinates or generates factually incorrect content.

“People have a very high degree of trust in ChatGPT, which is interesting, because AI hallucinates,” Altman stated during the discussion. “It should be the tech that you don't trust that much.”

Story continues below Advertisement

Altman’s remarks came amid growing public reliance on artificial intelligence tools, especially large language models (LLMs) such as ChatGPT, for tasks ranging from professional research and customer service to parenting advice and creative writing. The CEO himself admitted to using ChatGPT extensively for parenting guidance during his son’s early months, noting both its remarkable utility and its inherent risks.

The comment underscored what Altman described as a paradox at the heart of AI’s mainstream adoption: while hallucination—a known flaw in LLMs—remains a critical issue, users continue to trust and rely heavily on AI systems because of their conversational fluency, contextual memory, speed, and ease of use.