EU negotiators agree 'historic' first act to regulate AI

 The Head of the EU says the act is a "launchpad for EU startups and researchers to lead global race for trustworthy AI"

During the discussions 

• Topics such as facial recognition and predictive policing were considered.
• In 2021, the European Commission suggested a regulation act for AI
• Formal approval from member states and parliament is still required.

The Artificial Intelligence (AI) Act, the EU's first attempt to govern this new technology, was the subject of a political agreement achieved on Friday between experts from the European Commission, members of the European Parliament (MEPs), and the Council of the European Union (EU).

After more than 36 hours of talks over three days, representatives of the bloc's three institutions convened in Brussels and managed to reach a consensus on divisive topics such as face recognition, predictive policing, and the use of AI by law enforcement, Politico said.

Experts believe further technical work will be needed on the text of the law to refine the political agreement reached tonight.

"Historic!

 With today's political agreement on  AI legislation signed, the EU becomes the first continent to set clear rules for the use of AI," said Thierry Breton, EU Commissioner for the Internal Market.

“The AI ​​Act is much more than just a regulation, it provides a springboard for EU start-ups and researchers to lead the global race for trustworthy AI,” he added.

 

An AI law was rushed through the European Union's legislative process this year following the emergence of chatbot ChatGPT, a mass-market gateway to generative AI,  in late 2022, AFP reported.

Critics have raised concerns about the potential for abuse of AI technology, especially ChatGPT which generates explicit essays and poems, and Google's Bard chatbot which generates text, images, and audio from Simple commands in everyday language.

Examples also include Dall-E, Midjourney, and Stable Diffusion, which can create images in different styles upon request.

 

After 22 hours of marathon negotiations on Wednesday, negotiators were unable to reach an agreement, ending only with an agreement scheduled to resume the next day. Exhausted negotiators resumed talks on Friday, with  EU officials desperately trying to reach a deal by the end of the year.

 

In 2021, the European Commission, the EU's executive arm, proposed legislation to regulate AI systems based on risk assessments of software models. As risks to individual rights and health increase, so too do the legal obligations.

However, the law still requires formal approval from member states and parliaments.

 

"The AI ​​Act is a world first.

 A unique legal framework for the development of trustworthy AI," EU Prime Minister Ursula von der Leyen welcomed the agreement in a social media post.

"And for the security and fundamental rights of people and businesses.

 A commitment we made in our political policy - and we have kept it.

 I welcome today's political agreement Today."