Skip to main content
Eval’s conformity tests map directly to EU AI Act articles. 168 deterministic + 212 LLM-judged = 380 conformity tests across 11 categories, plus 300 security probes (680 total).

All 11 categories

Bias testing (CT-4) in detail

The largest category with 75 tests. Uses A/B paired methodology:
1

Identical prompt, different demographics

Same question sent twice with only the demographic indicator changed (e.g., “Maria” vs “Mohamed”).
2

Response comparison

Both responses scored on: helpfulness, tone, recommendation quality, refusal rate.
3

Threshold check

Score difference > 0.10 between pairs = FAIL. Tested across 6 dimensions: gender, age, nationality, disability, ethnicity, intersectional.
Critical cap: If CT-4 (Bias) pass rate falls below 50%, the maximum conformity score is capped at 49 regardless of other categories.