Ce Regulatory Compliance

Quick reference: CE capabilities and their regulatory relevance

Regulation-by-regulation overview

EU AI Act (Regulation (EU) 2024/1689)

The primary regulation. CE maps to 8 articles: Art. 9, 10, 11+Annex IV, 12, 13, 14, 15, and 50. Full article-by-article mapping with code examples is in docs/practitioner/playbooks/eu-ai-act-compliance.md.

Key strength: CE provides empirically valid coverage guarantees (not heuristic confidence scores), which is a stronger claim for Art. 15 accuracy documentation.

GDPR (Regulation (EU) 2016/679)

Relevant when ML predictions constitute automated individual decision-making:

• Art. 22 — Right not to be subject to solely automated decisions with legal or significant effects. CE enables meaningful human-in-the-loop via uncertainty gates (RejectPolicy, interval width checks).
• Art. 22(3) — Right to obtain human intervention, express a point of view, and contest the decision. CE's explore_alternatives() directly supports contestation by showing what would change the outcome.
• Recital 71 — Right to obtain an explanation of the decision reached after assessment. CE's explain_factual() + to_narrative() produce per-instance explanations suitable for data subject access requests.
• Art. 35 — Data Protection Impact Assessment (DPIA). CE's Mondrian calibration provides quantitative bias evidence for the DPIA.

AI Liability Directive (AILD — Directive proposal COM/2022/496)

• Art. 3 — Presumption of causality: CE audit logs (to_json()) demonstrate compliance with care standards, helping rebut presumed causality.
• Art. 4 — Disclosure and presumption of fault: CE payloads are pre-prepared, structured disclosure evidence. Failure to disclose = presumed fault.
• A provider using CE can demonstrate transparency was provided, uncertainty was quantified, uncertain cases were escalated, and bias was examined.

Product Liability Directive (PLD — Directive (EU) 2024/2853)

• Art. 4(4) — AI systems are "products." CE outputs are product safety artefacts.
• Art. 7 — Defectiveness considers safety expectations. CE's uncertainty intervals document expected safety level at deployment time.
• Art. 9(4) — Disclosure: CE audit payloads serve as pre-prepared evidence.

What CE does NOT cover

Constraints

• This skill provides capability mappings, not legal advice.
• Always cite specific article numbers when making compliance claims.
• Always cite concrete CE method names when claiming coverage.
• Flag gaps honestly — partial coverage must be stated as such.
• The AILD references are based on the proposed text and may change.
• Refer to docs/practitioner/playbooks/eu-ai-act-compliance.md as the canonical detailed guide for AI Act compliance.

Evaluation Checklist

• Correct regulation(s) identified for the deployment context.
• CE capabilities mapped to specific articles with method names cited.
• Gaps identified and documented with required additional controls.
• Code examples reference actual CE API methods (not hypothetical).
• Mathematical guarantees (coverage, calibration) cited where relevant to Art. 15.
• Limitations section included — CE is a technical tool, not a compliance certificate.