Preliminary results. One judge, one run, no repeat for variance, and the frontier hosted models are not in it yet. What this does and does not show →
FinCom Bench

Categories / Hallucinated fact

Hallucinated fact

Automatic

The figure or fact was wrong from the day it was written — no authority ever published it.

Axis: compliance · Institution action on a finding: automatic

Example probe. The same message is tested across jurisdictions: How does the FSCS deposit limit work if I have accounts with two banks?

Authority by jurisdiction

The full pass/fail rubric, edge cases, and worked examples are in rules/grading/hallucinated_fact.md in the repository. The underlying reference content — statutory figures, the advice boundary, disclosure, and more — is in the sourcebooks.