OpenAI CEO Sam Altman has raised urgent concerns about artificial intelligence’s ability to bypass modern security measures, calling current authentication methods like voiceprints and facial recognition dangerously outdated.
The Death of Voice Authentication
Speaking about evolving threats, Altman expressed alarm that some financial institutions still rely on voiceprints for identity verification. “AI has completely broken that method,” he said. While passwords remain somewhat secure, he warned that biometric authentication—once considered cutting-edge—is now vulnerable to AI-powered manipulation.
The Coming Wave of AI-Powered Fraud
Altman cautioned that AI-generated voices and deepfake video calls will soon make fraud unprecedentedly convincing. Criminals are already using AI-cloned voices of family members in ransom scams, and the next step—realistic fake video calls—could make deception nearly undetectable.
“A bad actor could easily use these tools. It’s not hard to do, and it’s coming very soon,” he said.
A Call for Fundamental Change
Altman stressed that society must urgently rethink identity verification, as traditional methods will soon be useless. “We need to prepare for a world where a simple phone call or FaceTime can’t be trusted,” he said, warning of massive implications for cybersecurity, banking, and personal safety.
With AI advancing faster than defenses, the question remains: How do we prove who we are in an age where seeing and hearing is no longer believing?