The European Union will be the first region in the world to fully regulate the uses of artificial intelligence (AI). The States and the European Parliament have reached, at midnight from Friday to Saturday, and after three days of intense and tough negotiations, a provisional agreement, a final text that must still be ratified by the two parties before it comes into force. , expected at the end of 2026. This text defines the obligations and rules by which a technology that is here to stay and that is completely transforming daily life, but that entails as many possibilities as risks, many of them not even yet imaginable, must be governed. .
The Twenty-seven assure that they want to guarantee the first and limit the second as much as possible, for which they have negotiated a “future-proof” law that contains sufficient flexibility to be able to regulate functions or technologies currently unknown, or to adapt to the changes experienced by the existing ones, as explained by sources familiar with the negotiations before those responsible for the agreement began their marathon closed-door meeting in Brussels on Wednesday. But the devil, as always, is in the details, and the “passionate” negotiations, as witnesses of the endless tug-of-war described them, were prolonged due to the fight between States and MEPs – and sometimes between each party against each other – over what Is it a risk or not and what exceptions and safeguards should be put in place to guarantee that individual fundamental rights are not put at risk in order not to harm the economy or the interests of the States.
“The EU becomes the first continent to establish clear rules for the use of AI,” celebrated the Commissioner for the Internal Market, Thierry Breton, one of the main promoters of the regulations. As he has tweeted, this is “much more than a package of rules, it is a shuttle for the startups “European companies and researchers lead the global race for AI.” “The EU is very proud to be the first major region to achieve such comprehensive regulation. The President of the Commission herself, Ursula von der Leyen, has also celebrated a “pioneering in the world” regulation, a “unique legal framework for the development of an AI that can be trusted”, she has greeted on X (former Twitter ) the head of the European Executive.
The negotiations have been truly marathon even by European standards. It started on Wednesday at three in the afternoon, local time, and lasted almost 40 hours, in one of the longest trilogues remembered in Brussels, according to the participants themselves. And, although in recent days all parties had indicated that there had been enough progress to be “cautiously optimistic,” the negotiators arrived at Wednesday’s first meeting with numerous issues still to be agreed upon, up to 22 points.
Among them, the thorniest were, as expected, the question of how to regulate the foundational models on which popular tools such as ChatGPT are based, as well as biometric surveillance systems (such as facial recognition), which in the end ended up being the most arduously negotiated point and prolonged the discussions until this Friday, due to the strong suspicions aroused by these technologies that potentially – and in some cases already actually, as seen in some countries with little democratic muscle – allow state supervision and control that can directly collide with the fundamental rights of citizens.
Finally, the failure of the negotiations has been avoided. In fact, the question of the founding models was resolved in the early hours of Wednesday to Thursday and what prolonged the discussions was the other red line for parliamentarians: practical uses of AI in surveillance through biometric systems such as facial recognition.
The European Parliament arrived at the meetings with a long list of practices crossed out of the law, such as the regulation of real-time biometric surveillance in public spaces, emotion recognition, predictive policing or biometric categorization. The States fought hard so that some of these practices could be carried out, at least by the forces of order, although for very few cases and with strong safeguards, and the negotiations extended longer than anyone had anticipated to try to bring positions closer. which, at times, seemed irreconcilable, although now, visibly, it has been achieved.
No other country still has regulation as complete as the European one. The president of the United States, Joe Biden, signed a decree in October that requires technology companies to notify the Government of any advance that poses a “serious risk to national security.” Days later, the British Prime Minister, Rishi Sunak, called a summit from which the first commitment of 28 countries and the EU on these systems emerged (Bletchley Declaration) and the creation of a group of experts to monitor their progress.
You can follow EL PAÍS Technology in Facebook and x or sign up here to receive our weekly newsletter.
#approves #worlds #artificial #intelligence #law