Overview

• Altman told Federal Reserve leaders that AI voice-mimicking models can defeat banks' decade-old voiceprint authentication and authorize fraudulent transfers.
• He warned that generative AI will soon enable deepfake video calls indistinguishable from customers, escalating security risks beyond voice alone.
• Altman said malicious actors could imminently release powerful cloning tools, making a global fraud crisis in banking a near-term threat.
• He cautioned that advanced AI virtual agents are poised to automate customer service entirely, delivering faster, error-free support without human agents.
• Federal Reserve Vice Chair Michelle Bowman has expressed willingness to collaborate with OpenAI on developing AI regulations and security standards.