EU AI Act Guidance Will Start With the Question: What Is AI?

For the European officials in charge of implementing the EU’s sweeping new AI law, one of the first big questions they must answer is how to define artificial intelligence.

EU policymakers recently approved a law that regulates use of the technology based on risk—with greater scrutiny for uses more likely to harm individuals.

The AI Act’s first provisions, set to come into effect early next year, will ban uses of the technology deemed unacceptably risky, including biometric characterization systems that infer characteristics like a person’s religious beliefs, political opinions, or sexual orientation.

A newly formed EU AI Office—part of the ...