technologyUpdated Aug 30, 17:08 EDT
AI governance crisis
Tracking: AI governance crisis
Current probability
53%
7-day move
→ -1.8pp
30-day move
+29.5pp
Confidence interval
51–54%
Evidence count
4,269
Probability history · last 90 days90 daily points · 4,269 total nudges
Days without new evidence carry forward the prior day’s posterior. Each visible inflection is a Bayesian update triggered by ingested evidence — see the stream below for per-nudge reasoning.
Evidence stream?Evidence streamEach row is one dispatch scored against this threat. Direction + magnitude = how the Bayesian update moved the probability. Reasoning is what the AI engine saw in the article that justified the nudge — auditable, not a black box.latest 0 of 4,269
No recent evidence nudges. New articles may not yet have been scored against this threat.
Prior · Posterior
Prior
20%
base rate before evidence
→
Posterior
53%
after 4,269 pieces of evidence
Methodology
This threat’s probability started at the prior — a base rate from historical data and expert elicitation. Every dispatch scored against the threat nudges the probability via Bayesian update. The visible chart is the posterior evolving with each piece of evidence. Confidence interval tightens as evidence count grows. Read the full approach at /about/methodology.