The EU’s AI Rules Are Now Live. Here’s How the First Real Test Is Going

European regulators have moved further and faster than most of the world in writing binding rules for AI systems, and companies operating across the continent are now navigating the first real-world test of a comprehensive AI regulatory framework.

The European Union’s comprehensive AI regulation, which categorizes AI systems by risk level and imposes correspondingly stricter requirements on higher-risk applications, has moved into active implementation, giving the world its first real test of a binding, continent-wide AI regulatory framework rather than the more voluntary guidelines many other jurisdictions have relied on so far.

Companies operating AI systems in higher-risk categories, including certain applications in hiring, credit scoring, and law enforcement, now face specific documentation, testing, and transparency requirements considerably more prescriptive than the general-purpose guidelines that had previously governed AI deployment in most markets.

Compliance costs are a genuine competitive concern for smaller companies

Smaller companies and startups have raised concerns that the compliance burden associated with the new rules falls disproportionately on organizations without the dedicated legal and technical staff that larger companies can allocate to regulatory compliance, potentially entrenching the market position of already-dominant AI providers.

“The rules apply the same requirements regardless of company size. In practice, a large company absorbs that compliance cost far more easily than a ten-person startup.”

With other jurisdictions watching the EU’s implementation closely as a potential template for their own future AI regulation, how smoothly the framework functions in practice, including whether the compliance burden concerns prove justified, is likely to shape AI regulatory approaches well beyond Europe’s own borders.