BRUSSELS 

EU lawmakers on Thursday endorsed a draft of a new law regulating Artificial Intelligence systems.

The EU institutions announced in a statement that the Internal Market Committee and the Civil Liberties Committee approved the draft for the European Parliament’s position for negotiations on AI rules.

If adopted, the legislation will be the first one in the world regulating AI systems, and it will ensure that the use of AI is “overseen by people” in a safe, transparent, and non-discriminatory way, the statement explained.

The draft requires foundation models, such as language processing chatbot ChatGPT, to respect additional transparency and copyright rules, and show that the content was generated by AI.

It “strictly prohibits” systems from using manipulative techniques, social scoring, or exploiting people’s vulnerabilities.

It also bans most forms of biometric surveillance, only granting exceptions for law enforcement and prosecution authorities in certain cases.

At the same time, even law agencies will be banned to use emotion recognition and predicting policing systems that exercise profiling based on criminal past or location.

After the approval of the European Parliament’s plenary session, the mandate will guide the European Parliament's representatives during the talks with EU member states and the European Commission.

Once they agree on the final law, the EP will sign it officially.