Brussels Wrote Rules. Washington Wrote a Playbook.

The EU and the US looked at AI risk and chose opposite tools.

Brussels wrote binding law. The EU AI Act sorts systems into tiers, assigns duties, and backs it all with fines. Break the rules, pay the price. That's the European way: clear lines, real consequences.

Washington wrote a playbook. The NIST AI Risk Management Framework, published in January 2023, is voluntary. It gives you four functions, govern, map, measure, and manage, and trusts you to use them. No fines, no tiers. That's the American way: guidance, not mandates.

Here's what experienced teams do: they use the playbook to follow the law. NIST's functions map neatly onto the Act's homework. Manage your risks, document your work, test your systems, keep humans in charge. Different packaging, same discipline.

Rules tell you what you must do. Playbooks tell you how. You need both.

I explain frameworks; I don't give legal counsel.