Article 10 — Data and data governance
APPLIES DEC 2, 2027 WAS 2026-08-02 AMENDED BY OMNIBUS
Chapter III · High-risk tier
Source: EUR-Lex · CELEX 32024R1689
Verified 2026-07-16
Source: EUR-Lex · CELEX 32024R1689
Verified 2026-07-16
PLAIN-LANGUAGE SUMMARY — ORIGINAL INTERPRETATION, NOT THE OFFICIAL TEXT
Where high-risk AI systems are trained on data, the training, validation and testing sets must meet governance standards fitting their purpose. Providers examine sources, representativeness and possible bias, and act to detect and reduce bias that could harm people or cause unlawful discrimination. Data must be relevant and, as far as possible, accurate and complete. Special-category data may be used for bias correction only under strict safeguards. Where no model training is involved, only testing data are covered.
AMENDED BY THE DIGITAL OMNIBUS ON AI
This provision's application date moved under the Omnibus. It now applies from 2027-12-02 (originally 2026-08-02). Source: Omnibus final text (OJ pending) · verified 2026-07-16
AT A GLANCE
Chapter: III — High-risk AI systems
Applies: Dec 2, 2027
Tier: High-risk tier
PENALTY EXPOSURE
€15M or 3%
of worldwide turnover — see Art. 99 penalties framework.
ELSEWHERE IN CHAPTER III — High-risk AI systems