Digital, Entertainment & Sports

Entry into force of the European Union Artificial Intelligence Regulation.

Carlos Lalanda ·

The European Union Artificial Intelligence Regulation enters into force!

The European Union Artificial Intelligence Regulation entered into force on 1 August. The final text was published on 12 July, following more than three years of work since it was proposed by the European Commission.

With the aim of protecting the health, safety and fundamental rights of individuals in the EU, including privacy, non-discrimination and freedom to conduct a business, this Regulation regulates trade in and the use of Artificial Intelligence systems. Reliable, transparent, human-centred and responsible AI technology is necessary to achieve sound technological innovation.

The Regulation is based on a robust categorisation of risks which determines the legal obligations of AI providers, importers and deployers whose activities have an impact within European territory.

Application will be progressive; the key dates for operators are as follows:

  • The first rules on prohibited AI practices will apply from 2 February 2025.
  • The rules on general-purpose AI models, governance and penalties will apply from 2 August 2025.
  • The key sections on AI systems classified as “high-risk” will apply from 2 August 2026.
  • Lastly, it should be noted that the measures on AI-powered safety components will apply from 2 August 2027.

To complement the effective application of the legislation, the competent national authorities of each Member State will be particularly relevant. Spain has taken an initial step forward with the recent establishment of the Spanish Agency for the Supervision of Artificial Intelligence (AESIA). (Link to previous LinkedIn post?) Other elements, such as codes of practice, conformity assessments for HRAIS, working groups and regulatory sandboxes , will be fundamental to the implementation of the Regulation that has just entered into force.

Next steps:

We will be monitoring how AI regulation, which has so far focused on risk prevention, is complemented by the AI liability framework proposed by the Union. We await the adoption of the AI Liability Directive, proposed by the Commission in September 2022, and the revision of the Directive on liability for defective products (PLD 85/374/EEC) and the legislation on machinery.

More publications