OpenAI CEO Sam Altman warned Tuesday that advances in AI could precipitate a global “fraud crisis” by enabling attackers to easily impersonate individuals, undermining existing security measures.
Speaking at a Federal Reserve forum on AI’s economic and societal impacts, Altman highlighted the vulnerability of voice‑print authentication systems, which some financial institutions still use to authorise high‑value transactions. “AI has fully defeated most of the ways that people authenticate currently other than passwords,” he said.
AI‑Enabled Impersonation
Altman’s remarks underscored alarming incidents in which AI‑cloned voices have been used to swindle families and infiltrate high‑level communications. The FBI has issued similar cautions about AI‑powered voice and video “deepfakes,” and US officials recently reported attempts to mimic US Senator Marco Rubio’s voice in calls to foreign ministers and state governors.
“Right now, it’s a voice call; soon it’s going to be a video or FaceTime that’s indistinguishable from reality,” Altman observed, noting that such capabilities demand new verification tools.
In response, Altman endorsed The Orb, a “proof of human” authentication system developed by non‑profit Tools for Humanity, designed to help distinguish genuine users from AI‑generated forgeries. He also warned of broader risks, cautioning that adversaries could one day harness so‑called AI “superintelligence” to orchestrate sophisticated attacks on critical infrastructure, ranging from power grids to bioweapons development, before defenses can adapt.
OpenAI Deepens Policy and Research
Altman’s presentation coincided with OpenAI’s announcement that it will open its first Washington, DC office early next year. The 30‑person team, led by Chan Park (Head of Global Affairs, US & Canada) and Joe Larson (Vice President of Government, formerly of Anduril), will host policymakers, preview emerging AI technology and offer training sessions for educators and government officials.
The office will also house research teams studying AI’s economic impact and strategies for expanding equitable access to the technology.
The timing aligns with the White House’s imminent release of its “AI Action Plan,” a policy framework aimed at balancing innovation with regulation to maintain US leadership in AI. OpenAI provided recommendations for the plan and has intensified its Capitol Hill engagement in recent months.
Impact on Jobs and Economy
Despite predictions from industry peers about widespread job displacement, Altman remained circumspect on AI’s labour market effects. “No one knows what happens next,” he said, acknowledging that while some roles will vanish, entirely new forms of work and “silly status games” may emerge as AI automates traditional tasks.
Complementing Altman’s speech, OpenAI economist Ronnie Chatterji released a report comparing ChatGPT’s productivity gains to past paradigm shifts like electricity and the transistor. The study notes that 20% of US users employ ChatGPT as a “personalised tutor,” with potential long‑term economic benefits for the predominantly 18–34 demographic.
Chatterji will lead further workforce impact research from the new DC office, collaborating with economists Jason Furman and Michael Strain over the coming year.
