Anthropic and OpenAI Push to Pace Frontier AI After Agent-Safety Incidents

Anthropic CEO Dario Amodei has called for frontier AI labs to slow the rate of capability gains, while OpenAI has separately argued that continued scaling should depend on confidence in safety. The debate turns recent agent-security incidents and recursive self-improvement data into a concrete test of evaluator access, lab coordination and whether voluntary commitments can change release incentives.