The EU AI Act is the EU legislation that came into force in August 2024 and is aimed at regulating the evolving AI-driven landscape on the territory of the EU.
Key Aspects of the EU AI Act
Even though the provisions of the EU AI Act also set certain requirements for the developers and deployers of the so-to-speak general-purpose AI tech (GPAI), the key obligations are being put on the high-risk AI system providers and encompass critical spheres of AI use, including, but not limited to, law enforcement, migration, critical infrastructure, remote biometric identification, education, etc.
In this respect, some of the key requirements include:
- risk management & quality management system development;
- technical documentation prep;
- ongoing data governance;
- detailed record keeping, and more.
In the separate article (Article 5), the EU AI Act also establishes a thorough list of prohibited practices, regarding the use of AI, from deployment of deceitful and/or manipulative activities to exploitation of potential End User vulnerabilities, like age or disability, development of social scoring systems, and more.
The EU AI Act Implementation Timeline
Even though the EU AI Act initially came into force back in August 2024, as of February 2025 only several of its provisions, like those regarding prohibited AI tech systems, have become applicable, with more specific regulations scheduled to take effect in August 2025 and beyond, up until August 2027.
Namely, it’s expected that by August 2025, the EU member states will form designated authority bodies for monitoring and assessment of the existing AI practices, as well as define particular fines, penalties and other non-monetary measures, if any, which will be applied in case of the AI Act infringement.
What is the EU AI Pact?
Aiming to encourage businesses’ proactive compliance with responsible AI practices, established by the EU AI Act, in 2024 the European Commission also introduced the so-called EU AI Pact, inviting companies to sign voluntary pledges, committing to practical actions in at least three of the following key aspects:
- AI governance strategy development & implementation;
- high-risk AI tech system mapping;
- AI awareness & literacy promotion.
As of Q2 2025, 100+ businesses, like Atlassian, OpenAI, Google, Snap, etc. have already signed the EU AI Pact, with more signatories expected to join it later this year, too.