The European Parliament has approved the regulation of the use of artificial intelligence (AI) systems across the EU, paving the way for the EU's Artificial Intelligence Act, which is the first of its kind globally.

The legislation aims to protect basic civil rights and protect against AI threats to society while fostering technology innovation. The new measures classify AI systems according to four levels of risk, ranging from minimal to unacceptable. Apps that deal with subliminal manipulation of vulnerable people, video games, and spam filters fall into the lower levels of risk. At the same time, social rating systems, predictive police tools, and remote facial recognition and biometric identification in public are assessed as the most unacceptable level of risk.

Violations of the new AI laws will be fined up to 30 million euros ($33 million) or 6% of a company's annual global revenue. The legislation will likely come into force in 2025, with member states responsible for enforcing the rules. Lawmakers hope that powerful AI developers will see the rules as a constructive framework for further responsible technology development in the EU's single market. The EU plans to cooperate with its American counterparts to develop a voluntary code of conduct.

View Update