Most high-risk AI systems (employment, education, essential services) must meet AI Act obligations from this date.
The EU AI Act's High-Risk AI System provisions become fully enforceable from August 2026 for all new systems and from 2027 for systems already in market. High-risk categories under Annex III include AI used in: education and vocational training (admissions, assessment), employment and workforce management (recruitment screening, performance evaluation), access to essential services (credit, insurance, healthcare), law enforcement (risk assessment, facial recognition), migration and border control, and administration of justice.
For APAC enterprises selling into EU markets — particularly APAC fintech and HR-tech companies deploying risk-scoring or talent-assessment AI — this deadline triggers mandatory conformity assessments, technical documentation requirements under Article 11, EU database registration, and post-market monitoring obligations. Companies who miss this window face market exclusion from the EU AI database and potential fines of up to €15 million or 3% of global turnover.
The most common compliance gap we see in APAC-headquartered AI vendors is insufficient technical documentation for model inputs, outputs, accuracy benchmarks, and bias assessments — documentation that EU notified bodies require as a baseline for conformity assessment. AIMenta's EU AI Act preparation service helps APAC software and AI vendors document their models to Article 13 and Annex IV standards, positioning them for EU market access before their EU-headquartered competitors.