Altman Sounds Alarm on Looming AI-Fueled Fraud Crisis

OpenAI CEO Sam Altman has issued a stark warning about a fast-approaching “fraud crisis,” raising concerns that artificial intelligence could soon make it nearly impossible to tell real people from digital impersonations. Speaking at a Federal Reserve event, Altman said AI has already undermined many identity verification methods and cautioned that the world is ill-prepared for what’s coming next.

“AI has fully defeated most of the ways people authenticate their identity today, except passwords,” Altman told an audience that included top figures from major U.S. financial institutions. He singled out the continued use of voice authentication by some banks as particularly troubling, calling it “a crazy thing to still be doing.”

His remarks come amid growing anxiety about AI-generated scams. In recent months, there have been multiple reports of criminals using AI to mimic the voices of children to defraud parents, and even a deepfake of Secretary of State Marco Rubio was used to contact officials, according to U.S. authorities.

Altman warned that voice impersonation is just the beginning. “Soon it’s going to be video or FaceTime that’s indistinguishable from reality,” he said, noting that while OpenAI isn’t actively developing these tools, the global community must act quickly to confront their misuse.

He is backing a project called The Orb by Tools for Humanity – a biometric identity system designed to help confirm real human presence in a digital world increasingly populated by AI.

Altman’s appearance at the Fed comes just as the White House is preparing to unveil its “AI Action Plan,” which will outline the federal government’s approach to regulating the rapidly advancing technology. OpenAI has provided input on the plan and is increasing its presence in Washington, D.C., where it plans to open its first office in early 2026.

The new D.C. office will be led by Chan Park, OpenAI’s head of global affairs for the U.S. and Canada, alongside incoming vice president of government, Joe Larson. It will serve as a hub for policy engagement, AI education programs, and ongoing research into the socioeconomic effects of the technology.

Despite raising the alarm on AI-related fraud, OpenAI is also lobbying against overregulation. The company has cautioned that restrictive legislation could hamper innovation and weaken U.S. competitiveness in the AI race – particularly against China. Earlier this month, lawmakers removed a controversial Trump-era proposal that would have blocked states from introducing their own AI laws for a decade.

Beyond fraud, Altman spoke about the broader risks of superintelligent AI systems falling into the wrong hands. He voiced concern over scenarios where adversarial nations could use AI to disrupt power grids or develop bioweapons. He also warned of the dangers of handing too much decision-making power to future AI models, especially those that surpass human intelligence – a goal that several tech firms, including OpenAI, are openly pursuing.

On the issue of AI and jobs, Altman took a more neutral stance. While some industry leaders have warned that AI will eliminate vast numbers of roles, Altman said the future is far too uncertain to predict. “No one knows what happens next,” he said, describing most predictions about AI’s economic effects as “smart-sounding guesses.”

He acknowledged that some job categories will disappear, but expects new roles to emerge. Looking further ahead, Altman even speculated that traditional employment may become obsolete altogether. “You have everything you could possibly need. You have nothing to do,” he said of a hypothetical future. “So, you’re making up a job to play a silly status game and to feel useful to other people.”

To support this vision, OpenAI released a new report alongside Altman’s speech. Written by its chief economist, Ronnie Chatterji – a former White House adviser – the report compares AI’s potential to that of electricity and the transistor. It claims ChatGPT now has 500 million users globally, with 20% of U.S. users treating it as a personalized tutor. More than half of its American user base is under 35, which the report suggests could offer long-term economic advantages.

Over the next year, Chatterji will partner with economists Jason Furman and Michael Strain on a deeper study into AI’s effects on employment, much of which will take place at OpenAI’s D.C. office.

As Altman made clear during his remarks, the road ahead is filled with both extraordinary opportunity and enormous risk. For now, he believes the world is only beginning to grapple with the implications.