Over the coming months, our new series is going to explore the challenges and opportunities presented by AI. In our latest video, we're focussing on the basics of the EU AI Act. which recently came into force.

The EU AI Act is the EU's response to regulating AI, marking the first comprehensive AI regulation by a major regulator. It establishes a framework of obligations for entities making and using AI systems, depending on the risk level presented by its use case.

The Act will become fully enforceable in 2026, applying to providers, deployers, importers and others in the EU, with some provisions extending to entities outside the EU if their AI outputs are used within its jurisdiction. Notably, the Act does not cover military AI, national security applications or AI used solely for scientific research and development.

Additionally, the Act defines AI systems broadly, focusing on the ability to infer outputs from inputs. This definition has been criticised for its broadness, and the European Commission has clarified that traditional software systems are not included. Certain exclusions also exist, such as for low-risk open-source AI and AI used in personal, non-professional contexts.

For more information, take a look at our quick-fire round of questions on reedsmith.com.