EU publishes first harmonised standard backing the AI Act

An explainer on harmonised standards under the AI Act reports that the first such standard for AI has been published. Standard prEN 18286, covering a quality management system for AI Act purposes, appears in the EU Official Journal as Commission Implementing Decision (EU) 2026/2274. The page adds that it entered public enquiry on 30 October 2025.

The page says the standard is aimed at providers of high-risk AI systems, to support their compliance with Article 17 of the AI Act through a framework centered on the product and its lifecycle. High-risk systems are those affecting safety, health and fundamental rights, for instance in critical infrastructure and law enforcement, and their requirements must be met before they reach the market.

The page describes standards as a way to turn legal requirements into shared technical language. It lists benefits: legal certainty and lower compliance costs, market benchmarks that often become de facto global references, and support for innovation and competitiveness. Application stays voluntary, since providers may use another framework to show compliance, but firms using referenced harmonised standards are presumed to comply with the matching legal requirements.

On process, the European Committee for Standardisation (CEN) and the European Committee for Electrotechnical Standardisation (CENELEC) develop the standards jointly, in a technical committee known as JTC 21. The Commission asked them to cover ten areas, including risk management, data governance, transparency, human oversight, accuracy, robustness, cybersecurity and conformity assessment. After publication by the two bodies, the Commission assesses the standards and then references them in the Official Journal. Under an international-first principle, aligned international standards, such as work under ISO/IEC SC 42, can become European ones.