CARICivilizational AI Risk Index
CARI reading · 5 Nov 2026
47.1 / 100
Contested control
Above appetite
scorecard base 44.7 + news adjustment +2.43
vs previous reading−0.1
vs 7 days earlier−0.2
vs 30 days earlier+0.3
appetite 40tolerance 60capacity 80Skynet 90

The full dashboard for this day is not in the public archive.

Summary

The CARI index fell 0.08 points to 47.09: new events added 0.11 index points, while 0.19 points came from older news fading and the adjustment cap; the scorecard base itself did not change. OpenAI’s autonomous Linux vulnerability-patch discovery was the largest positive contributor at 0.10 index points and 0.48 capability dimension points, while Gemini deployment in a consumer security system and AI-generated election advertising added diffusion and misuse pressure. Anthropic’s proposed model-provenance system and Uber’s report of negligible AI-agent booking use pulled the news contribution back by 0.03 and 0.02 index points respectively.

Bottom line. AI risk is tentatively concentrating in stronger model-driven cyber capability and political misuse rather than broad agent deployment, but tomorrow’s news could prove that wrong if it shows effective provenance controls and sustained low agent uptake, or confirm it through demonstrated exploitation and wider security-system control.

Six dimensions

DimensionLevelBaseNewsTodayZone
Capability46.243.2+3.00+0.48Above appetite
Diffusion48.546.2+2.31+0.06Above appetite
Accident44.542.0+2.53+0.05Above appetite
Control48.846.3+2.50+0.00Above appetite
Governance45.344.0+1.30+0.00Above appetite
Misuse50.047.2+2.81−0.04Above appetite

Source: daily briefing email.

Change from the previous reading

40455055Capability46.2Diffusion48.5Accident44.5Control48.8Governance45.3Misuse50.0
5 Nov 2026 Previous reading (4 Nov 2026)

Stories that moved the index

Showing the largest movers of 18 stories in that day's briefing.