AI Act high-risk + LPP + GDPR. One signed pack.

Yes, in most judicial and access-to-justice contexts. Annex III section 6 / Article 6(3) capture AI used to assist a judicial authority in researching and interpreting facts and the law, plus AI affecting access to essential public services and law enforcement. That triggers the full 9-Article high-risk stack: risk management (Art 9), data governance (Art 10), technical documentation (Art 11), record-keeping (Art 12), transparency (Art 13), human oversight (Art 14), accuracy/robustness (Art 15), QMS (Art 17) and conformity assessment (Art 43).

LPP-protected communications must not flow through AI systems without explicit, informed client consent. MEOK enforces LPP boundaries at the data layer: privileged material is fenced from training, retrieval and logging pipelines, and the audit trail records that the boundary held. Breaking privilege is not just a compliance failure, it can be a regulatory and negligence liability for the firm.

GDPR Article 6 requires a lawful basis for any AI processing of personal data, and Article 22 gives individuals the right not to be subject to solely-automated decisions producing legal or similarly significant effects. In sentencing, bail, asylum and similar contexts that means human-in-the-loop review must be substantive, not a rubber stamp. A DPIA is expected wherever the processing is high-risk.

AI used in judicial decisions, sentencing, bail and asylum can produce disparate impact across protected groups. MEOK ships demographic-parity and equalized-odds testing aligned to ISO/IEC 42005, satisfying EU AI Act Article 10 data-governance and bias-examination duties and giving procurement teams signed evidence that ranking and scoring outputs were fairness-tested.