The EU AI Act entered into force on 1 August 2024, starting a phased rollout across the European Union. Six months later, on 2 February 2025, the first major set of rules took effect. This milestone banned all prohibited AI practices and activated Article 4 AI literacy obligations, which require all AI providers and deployers to ensure their staff understand AI risks and compliance duties.

One year after entry into force, on 2 August 2025, rules governing General Purpose AI models became active. This milestone introduced governance rules for model providers, mandatory technical documentation, copyright policy compliance, and oversight authority for the European AI Office.

On 2 August 2026, the AI Act reached general application. This key milestone activated Article 50 transparency duties for chatbots, deepfakes, and synthetic content, alongside national authority enforcement and penalty frameworks. As of August 2026, these general application and transparency rules are fully live and enforceable across all Member States.