EU Committee Approves First AI Legislation

European Parliament accepted a groundbreaking legislation for AI proposed by the EU with a vote scheduled in April.

Parliamentarians in the European Parliament accepted the preliminary agreement on ground-breaking legislation for artificial intelligence (AI) on February 13.

This agreement was brought about by the European Union. A vote in the parliament is slated to take place in April, which thus paves the way for the world’s first legislation that focuses on artificial intelligence.

The Internal Market and Civil Liberties Committees approved the interim agreement on the Artificial Intelligence Act with a vote of 71-8. The purpose of the act is to establish criteria for artificial intelligence across a variety of industries including banking, automotive, electronics, aviation, law enforcement and security operations.

The fundamental models, or generative artificial intelligence, will be supervised by the legislation. OpenAI’s ChatGPT trains these models on extremely large data sets. For the purpose of addressing the issue of generative artificial intelligence models, the AI Act intends to implement protections including copyright protection for inventors.

In addition, it restricts applications of artificial intelligence that represent a risk to the rights of citizens such as social grading and biometric categorization. France retracted its objection which led to concessions aimed at decreasing the administrative load on high-risk AI systems and giving better protection for commercial secrets.

The endorsement comes after acceptance by EU member states, which followed the withdrawal of France’s objection. After reaching a political agreement in December 2023, parliamentarians attempted to translate the agreed-upon positions into a final compromise document for approval.

The permanent representatives of all member states held the “coreper” vote on February 2. The Committee on Civil Liberties of the European Parliament published an article on the social network X, pointing to the endorsement as an example of artificial intelligence (AI) making progress.In either March or April, the Artificial Intelligence Act is scheduled to be put up for a vote in the European Parliament.

If approved, the full implementation is anticipated to occur twenty-four months after it enters into force with certain clauses taking effect earlier.The regulators received a letter from a consortium of enterprises and technology companies in November 2023, which warned them against excessively monitoring artificial intelligence (AI) systems that are robust at the expense of innovation.

Thirty-three businesses that are active in the European Union (EU) signed the letter which underlined that excessively stringent laws on foundation models and general-purpose artificial intelligence could potentially impede critical innovation in the region.

The European Commission is now establishing an artificial intelligence office to ensure adherence to a handful of high-impact foundational models that are thought to pose systemic concerns. In addition, it revealed plans to provide assistance to local artificial intelligence developers, such as by enhancing the European Union’s supercomputer network for the purpose of training generative AI models.