The European Parliament has approved large-scale legislation on AI regulation. This happened almost three years after the draft rules were first proposed, Engadget writes.

This is the so-called Law on Artificial Intelligence. Last December, it was agreed upon in negotiations with EU member states. Now, it has been approved by MEPs. The document received 523 votes in favor and 46 against. Another 49 deputies abstained.

European legislation defines obligations regarding the use of artificial intelligence based on its potential risks and level of impact.

The law has not yet entered into force – it has yet to be reviewed by legal experts. In addition, the European Council must officially enact it.

The penalties for violating the law can be very severe. Companies that violate the new rules will be fined up to 35 million euros ($51.6) or up to 7% of their global annual revenue. It will depend on which of these amounts is higher.

The AI law applies to any model operating in the EU. Therefore, American companies operating in this field will have to comply with it, at least in Europe.

To ensure compliance with the law, each EU member state will set up its own supervisory authority, and the European Commission will establish an AI Office to develop methods for assessing and monitoring the risks of AI models.