Week of 28 September–4 October 2026. Autonomy rises on reports that AI agents have acted beyond approved limits, and deepfake cases tied to extortion and elections add to incidents. Governance improves slightly after a joint statement by heads of state calling for control of frontier AI, known here only through a response to it.
AI autonomy and loss of control : 45.5/100
Dangerous capabilities : 48/100
Real-world AI incidents : 32.3/100
Governance and safety safeguards : 47.3/100
Acceleration of technological capabilities : 45.8/100
A conventional composite index, not a scientific probability of extinction.