National legal AI risk dashboard
Source-backed observed patterns. Corpus checked Sep 10, 2026; latest tracked decision Sep 8, 2026. This dashboard does not estimate the prevalence of AI use or usage-adjusted incident rates.
Observed filing-risk patterns require a verification gate, not a blanket AI ban.
Prioritize citation existence, quotation accuracy, proposition support, responsible-reviewer signoff, and a matter-level exception record.
Editorial impact mix
Impact is an editorial consequence classification, not a probability or prediction.
Recorded matter attributes
Tags include posture, participants, failure patterns, and outcomes. They overlap, so one matter may contribute to several bars.
Recorded matters over time
Counts reflect the current corpus snapshot and may rise for earlier years as research coverage expands.
Source-link tier
A linked source improves traceability; it does not mean every field has been independently verified.
Explicit tool mentions
Mentions are not usage-adjusted failure rates and do not establish comparative product safety.
Recorded practice signals
Practice labels describe recorded matter context, not market-level exposure.