EU's AI 'guardrails' cannot absorb rapid changes in technology, study warns

July 17, 2026 - 07:06

The European Union's approach to regulating artificial intelligence is not working as intended, according to a new study published in the journal Big Data & Society. Researchers argue that the so-called "guardrails" the EU has built to govern AI are too weak in their goals and too clumsy in their execution. The rules, they say, have become so heavy and rigid that they cannot keep up with the speed of technological change.

The study points to a fundamental mismatch. The EU's AI Act, which was designed to classify systems by risk and impose strict requirements on high-risk applications, was written with a specific vision of how AI would evolve. But the technology has moved far faster than lawmakers anticipated. Generative AI models, for example, were not the main focus when the rules were first drafted. Now they dominate the conversation, and the existing framework struggles to address them.

The researchers warn that the EU's method of trying to predict and pre-empt risks through detailed legal definitions is failing. Instead of creating a flexible system that can adapt, the rules lock in assumptions that quickly become outdated. The study calls for a more dynamic approach, one that does not rely on static categories but instead builds in regular review and adjustment. Without that, the authors conclude, the EU's regulatory model risks becoming irrelevant just as AI's impact on society grows.