From fun to fallout: Sam Altman warns user on the surprising & risky ways people use ChatGPT

OpenAI CEO Sam Altman has issued a cautionary note to ChatGPT users: conversations with the AI chatbot are not legally protected or confidential, and anything shared could be subpoenaed in court. 

Speaking on comedian Theo Von’s podcast This Past Weekend, Altman reflected on the wide range of ways people are using ChatGPT — some of which are intensely personal or sensitive. From analyzing bank statements to writing Studio Ghibli fan fiction, Altman noted that users are pushing the boundaries of what the technology was originally built for. 

“Some people are having it read bank financial statements. Some people are writing Studio Ghibli fan fiction with it,” he said. “The spread is really wide.” 

While he celebrated this organic adoption, Altman expressed concern about the level of trust people place in the tool. Unlike conversations with doctors, lawyers, or therapists, chats with ChatGPT do not carry any form of legal privilege, making them potentially accessible in legal proceedings. 

“There are moments of real existential weirdness,” Altman said, reflecting on the weight of being at the center of such a powerful shift. “But most of the time it’s just really fun.” 

He emphasized that ChatGPT has evolved into more than a productivity tool—it’s become a creative partner and thinking companion for many. This influence, he said, isn’t neutral. “It probably has a big impact on what [people] believe and how they think,” Altman noted. “It’s not neutral, and we shouldn’t pretend it is.” 

Now a father, Altman described AI’s growth in emotional terms: “Beautiful and scary at the same time.” He believes AI’s potential to help people tell stories and understand themselves is something to be proud of. “People want to tell stories. They want to understand themselves. If AI can help with that — even just a little — that’s something to be proud of.” 

Looking to the future, Altman discussed how models like a potential GPT-7 could fundamentally reshape how we live and work, possibly prompting new conversations around productivity and economic redistribution — such as AI dividends. 

He also predicted a shift away from traditional apps and interfaces: “At some point, you’ll just tell your AI, ‘Book a flight to New York next Thursday’ — and it’ll go do it.” 

Still, he urged caution. Altman stressed the importance of transparency, better regulation, and awareness of AI’s environmental impact. “We try to be thoughtful, but this is going to have enormous impact. And we don’t fully understand it yet.” 

Despite the uncertainties, Altman remains optimistic about AI’s role in human expression. “It’s cool to build a tool that people find useful. And it’s even better when people find it meaningful.”