Sam Altman, CEO of OpenAI, has voiced serious concerns over the potential for artificial intelligence to fuel a global wave of fraud. Speaking at a recent event hosted by the Federal Reserve, Altman highlighted the increasing ability of AI to mimic human voices and identities—posing a serious threat to security systems that still rely on voice authentication.
“What really worries me is that some financial institutions still accept voiceprints for authorization,” Altman said. “That’s extremely risky in today’s AI-driven landscape.” He stressed that AI has already outsmarted many common verification methods, leaving passwords as one of the few remaining defenses.
Washington Presence & AI’s Expanding Role
Altman’s remarks come as OpenAI prepares to deepen its presence in the U.S. capital. The company is set to open its first Washington, D.C. office next year, where it will host policymakers, run educational sessions, and conduct research into AI’s economic impact and accessibility. Chan Park and Joe Larson will lead the office, focusing on both policy outreach and strategic government engagement.
This expansion aligns with growing government interest in AI governance. The White House is expected to unveil an “AI Action Plan” that outlines strategies for managing the opportunities and risks of AI development. OpenAI has contributed to the formulation of this plan and has encouraged balanced regulations that won’t hinder innovation.
Despite acknowledging the dangers, OpenAI remains cautious about overregulation, especially when it comes to maintaining the U.S.’s competitive edge in global AI development. A recent Senate decision rejected a proposed ten-year restriction on states creating their own AI laws—something OpenAI opposed, citing concerns over innovation stifling.
Fraud Risks and Deepfake Threats
Altman’s warnings are not without basis. U.S. authorities have seen a rise in AI-powered scams involving cloned voices and manipulated videos. There have been cases where scammers used fake audio of children crying to extort money from parents, and even attempts to impersonate high-ranking government officials through AI-generated audio.
Altman noted that these threats are only the beginning. “Soon it won’t just be voice. We’ll be dealing with video calls and digital impersonations so convincing they’re impossible to distinguish from reality,” he said. Although OpenAI is not developing such tools, the company is aware of the dangers and supports initiatives like The Orb, a project aimed at verifying human identity online.
Uncertain Job Landscape, But Optimism Persists
While some tech leaders warn that AI will drastically shrink the job market, Altman remains more cautious in his predictions. “No one really knows what will happen,” he said, adding that AI is too new and complex for accurate forecasting. He acknowledged that some job categories will disappear, but new roles will also emerge.
Altman even speculated that in a future dominated by AI, many people may not have traditional jobs at all. Instead, they might pursue tasks for fulfillment, creativity, or social interaction—rather than economic survival.
To study AI’s real impact on the workforce, OpenAI’s chief economist Ronnie Chatterji has released a report highlighting productivity gains from ChatGPT, which now boasts 500 million global users. A significant portion of young U.S. users leverage it for learning and skill development, potentially signaling long-term economic shifts.
Chatterji plans to collaborate with leading economists to conduct a more detailed study of AI’s role in reshaping labor and productivity in the coming year.
