Changelog

1. v1.1.03 August 2026

   Reclassify medical safety-component AI under Article 6(1) / Annex I.

   • Removed fabricated Annex III medical area that did not exist in the Act.
   • Medical diagnosis / device safety-component patterns now cite Article 6(1) and Annex I.
   • Applicable deadline for that path is 2028-08-02 (Annex I embedded), not the Annex III track.
   • Annex III point 5 remains access to and enjoyment of essential private and public services and benefits.

   Depends on lib/ai_act/enforcement-dates.ts for the Annex I calendar date. instrumentInForce from LEGAL_INSTRUMENTS reg-2026-1744.

   Source:eur-lex.europa.euRegulation (EU) 2026/1744 (Digital Omnibus on AI)
2. v1.0.017 July 2026

   Baseline EU AI Act rules corpus.

   • Article 5 prohibited practices (subliminal/manipulative, social scoring, real-time remote biometric ID, etc.).
   • Article 6(2) / Annex III high-risk domains: biometrics, employment, education, essential services, law enforcement, migration, justice.
   • Article 6(3) exceptions (narrow procedural task, improves human activity, pattern-detection, preparatory task).
   • Chapter V GPAI obligations, standard (Article 53) and systemic-risk (Article 55).
   • Article 50 transparency: chatbot disclosure and synthetic-content (deepfake) labelling.

   Enacted application dates apply. The Digital Omnibus (provisional) may defer Annex III high-risk and Annex I product dates; tracked here and applied in a future version once final.

   Source:eur-lex.europa.eu

Article references are to Regulation (EU) 2024/1689 (the EU AI Act). This is compliance information, not legal advice.