§8.4 · BUILDING · 5 MIN READ · EDITION 1.0
Seeded difficult cases and safe failure
A capability should be tested deliberately against cases it is expected to struggle with, to confirm it fails safely rather than confidently wrong.
Seed the reference set with cases chosen because they are hard: ambiguous wording, missing pages, conflicting revisions, borderline dimensions. A tool that reports low confidence or asks for escalation on these cases is behaving safely; one that reports a confident wrong answer is not.
Run this diagnostic on a recurring cadence, not only once at qualification, since evidence sources and model behaviour both change over time.
PUT THIS INTO PRACTICE
- 01Deliberately seed known-difficult cases into testing.
- 02Treat low-confidence or escalation responses as a pass, not a failure.
- 03Repeat the diagnostic on a defined cadence after go-live.
RELATED INSTRUMENTS
CITE THIS SECTION
§8.4 · Edition 1.0 · Seeded difficult cases and safe failure