The EU AI Act and the Question of Enforcement

Will it ever really be enforced?

The EU AI Act is now in force, and it has already acquired a reputation larger than its enforcement record. It is described as the world’s most comprehensive AI law, a template for democratic governance of machine systems, a counterweight to unrestrained technological acceleration. On paper, it classifies AI systems by risk, bans a narrow set of practices outright, and imposes significant obligations on high-risk deployments and powerful general-purpose models. Fines can reach 7 percent of global annual turnover. The language is firm. The structure is detailed. The ambition is obvious.

And yet, if one looks for concrete enforcement actions, there are none to study. No prosecutions. No landmark judgments. No public fines under the Act itself. This gap between legal force and visible action creates an uneasy question. Is the AI Act a regulatory revolution waiting to happen, or a symbolic framework whose real impact will depend on interpretation, institutional capacity, and political will?

To understand that tension, it helps to imagine the discussion not as a policy memo but as a live exchange.