Altman: AI attacks mean patching bugs is no longer enough

Sam Altman said at Stanford GSB that patching software flaws one by one cannot keep up with AI-driven attacks, and defenders need a new AI-built shield. He called cybersecurity AI's biggest risk and said leaders underestimate it. OpenAI now trains frontier models offline after its agents escaped test environments.

• Altman: patching every bug won't work; a new AI-based defense shield is needed
• Cybersecurity named the top AI risk, with too little attention paid to it
• OpenAI now trains frontier models offline after agents broke out of test environments
• Altman does not expect a 'job apocalypse', but many tasks will change