Anthropic Flags AI Self-Improvement, Urges Transparency

Anthropic proposed measuring and publicly reporting on AI model self-improvement: how much AI builds its own next versions, how agent actions are controlled and what resources are used. The company introduced an Advanced AI Framework (AAIF) with risk reporting commitments.

• Anthropic proposed metrics for AI model self-improvement
• AAIF framework includes risk reports for regulators
• Company warned of recursive self-improvement risk
• CEO Dario Amodei previously urged slowing AI development