OpenChainGraph Suite · ART-05 · EU AI Act Conformity
v1.0.0

EU AI Act Credit-Scoring Conformity Pack

Pre-assess your credit-scoring AI system against EU AI Act (Regulation (EU) 2024/1689) Annex III Part 5(b) high-risk obligations — fully applicable 2 December 2027. Enter your model's demographic performance metrics, answer the attestation checklist, and get a conformity determination with every failing dimension enumerated, plus a downloadable Article 11 documentation skeleton.

AP2 Export Chains: PTG-01 Deadline: Aug 2, 2026 EU AI Act Annex III 5(b) Zero PII Client-Side · No Inference
🔒 All inputs are processed locally in your browser. No data is transmitted. Do not enter real personal data — use synthetic or anonymised inputs only.
⏰ 2 December 2027 — EU AI Act Annex III Part 5(b) high-risk AI obligations for credit scoring and creditworthiness assessment fully apply. This includes Articles 9 (risk management), 10 (data governance), 11 (technical documentation), 13 (transparency), 14 (human oversight), and 15 (accuracy and robustness). Entities subject to this regulation must complete conformity assessment before this date.
⚠ Educational pre-assessment only. Bias metrics are computed on user-entered synthetic or anonymised performance data — results are indicative, not a substitute for a full third-party conformity assessment or notified body certification. Regulatory citations are real (EU AI Act, Regulation (EU) 2024/1689). The 0.8 disparate impact threshold is a widely used indicator (EEOC four-fifths rule, referenced in EU anti-discrimination guidance); the EU AI Act does not specify a numeric threshold — verify with legal counsel. Do not upload real personal data.
AI system details
Bias testing — protected characteristics

Enter your model's approval rates and error rates (TPR / FPR) per demographic group for each protected characteristic. Use synthetic or anonymised aggregate statistics — do not enter individual records. Disparate Impact Ratio (DIR) = lowest group approval rate ÷ highest group approval rate. DIR < 0.8 flags adverse impact. Equalized Odds Gap = max spread in TPR or FPR across groups — gap > 5pp flags imbalance.

Data governance attestations — Art. 10
Conformity checklist — Art. 9 / 11 / 13 / 14 / 15
Dimension scores
Bias metrics — per characteristic
Article 11 technical documentation skeleton
Chain to PTG-01 — generate regulatory submission draft
→ Open in PTG-01 Export artifact below, paste into PTG-01 → generate Article 11 submission draft or board memo