Europe's first applicable AI standard is ready, with Denmark in the lead role

Europe now has its first harmonised standard supporting the EU's AI Act, and Denmark led its development. The standard, EN 18286, covers quality management of AI systems. Dansk Standard, the Danish standards body, led the work and is also heading the development of several upcoming AI standards meant to support the regulation's implementation. It was harmonised in the EU's Official Journal on 14 October 2026.

EN 18286 gives companies a shared framework for quality, risk management, documentation and continuous improvement of AI systems. The AI Act sets out what companies must meet, while the standard describes how they can work with those requirements in practice: identifying risks, documenting AI solutions and building safety and accountability into new systems. The Act, adopted in 2024, places particular demands on high-risk systems, such as those used for credit scoring, recruitment and biometric identification.

A harmonised standard is a European standard developed at the European Commission's request to support specific EU legal requirements. Once published in the Official Journal, it gives a presumption of conformity. Organisations with high-risk AI systems can therefore comply with the law by following the standard when the rules take effect on 2 December 2027.

The standard is the result of cooperation among businesses, authorities, research institutions and experts across Europe. The work was complex because AI develops quickly and no existing standards were available to build on.

Dansk Standard's director, Torben Jansen, said Europe is unlikely to compete with the most open AI solutions by imposing the fewest requirements, but can lead in trustworthy and responsible AI, with trust potentially becoming a strong competitive advantage. Danish minister for research, education and digitalisation Christina Egelund said the standard strengthens companies' competitiveness and trust in continued work with AI.