The European Parliament has given its massive approval to the artificial intelligence regulation, which now only needs formal ratification by the States to become the first comprehensive law in the world that seeks to regulate a technology that is transforming the planet, not always for the better. This Wednesday, MEPs validated the regulations, which seek, according to their speakers, precisely to protect fundamental citizen rights without hindering innovation, by 523 votes in favor, 46 against and 49 abstentions.

The Commissioner for the Internal Market and the main promoter of the law, Thierry Breton, has welcomed the “overwhelming” support of MEPs for a law that, he stated, turns Europe into a “global regulator of reliable AI.” AI “is already part of our daily lives. Now, it will also be part of our legislation,” said the President of the European Parliament, Roberta Metsola, who led the vote in Strasbourg.

Breton, for his part, has assured that he is “ready” for the implementation of the regulation, which may begin to come into force in a few months: on the side of the European Parliament, there remains one last “legal-linguistic check” since the text of the law has not yet been able to be translated into all the official languages, something that will be done in principle without obstacles (especially in view of the strong support for the regulations in the key vote this Wednesday) in the next and last plenary session of this mandate, in April. The main thing now is for the States to also give it their final approval, something that according to diplomatic sources could happen around the month of May. Although this last step should be a mere procedure – since it was already approved unanimously by the representatives of the countries in February – the recent precedents at this stage, with countries – especially Germany, but also France or Italy – taking action backing down at the last minute in already negotiated agreements means that no one puts their hand in the fire until that last step is taken.

In any case, once approved, it will come into force 20 days after being published in the Official Gazette, at which time its gradual implementation will begin—being a regulation, it is applied directly without having to adapt it to national legislation—until its full application in 2026. The first steps will be the launch of the AI ​​Office, the body that will supervise compliance with the regulations, and which Breton, under whose portfolio it will operate, has announced that “it will promote the European AI strategy , that is, an AI that can be trusted, that respects European values ​​and standards and that is trusted by citizens and companies around the world.”

Although the EU already promotes a voluntary and gradual implementation of future rules for AI in companies and institutions, one of the parts that will come into force for everyone at the latest before the end of the year is the ban on prohibited intelligence systems artificial, which according to the regulations must be implemented six months after the law is approved.

An important moment, highlighted one of the rapporteurs of the regulation, the Italian social democrat Brando Benifei, because it will make it clear, he said at a press conference from Strasbourg, “that some cases are prohibited and are not welcome in Europe.” Codes of good practice must be in place nine months after the law is passed, general-purpose AI standards, including governance, 12 months later, and finally obligations for high-risk systems in 36 months.

The AI ​​law establishes different requirements and obligations for AI applications depending on the risks posed by their use. The most innocuous, such as filters spam or text duplicate detectors, can be used without any restrictions. They are called a limited risk system and the only requirement placed on providers is that they inform users that they are using an AI tool.

On the contrary, those considered to have unacceptable risk are completely prohibited, while those of high risk will require permanent supervision. The first category includes systems “that transcend a person’s consciousness or deliberately manipulative techniques”, those that exploit their vulnerabilities or those that infer people’s emotions, race or political opinions. For their part, remote biometric identification systems are considered high risk, which a large sector of Parliament wanted to strictly prohibit, although they had to partially yield to the States, biometric categorization systems or emotion recognition. Also systems that affect the security of critical infrastructure and those related to education (behavioral evaluation, admission and examination systems), employment (personnel selection) and the provision of essential public services, law enforcement or migration management.

Although the regulations do not go all the way in terms of guarantees against interference by the States or the technologies themselves in the freedoms and rights of citizens, Benifei has insisted that the law has sufficient “safeguards.” “We are convinced that with this text there is no risk of mass surveillance, because we have imposed extremely strict safeguards thanks to a very tough negotiation,” he stated.

One of the issues in the regulation that has attracted the most attention is how copyright is protected. Although the regulations ensure that all systems must guarantee compliance with copyright, it does not say how, other than that it will be the responsibility of the AI ​​Office. Despite this, some of the main associations that bring together creators and rights holders from the culture and press sector in Europe have welcomed the approval of the law in a joint statement as an “example of responsible governance of AI”, if They have called on the European Parliament to “continue support for responsible and sustainable AI by ensuring that these important rules are put into practice in a meaningful and efficient way.”

It has not been an easy road to get to this Wednesday’s vote in Strasbourg. The regulation has been under discussion for five years and its final text was only agreed last December, still under the Spanish presidency of the EU, after 38 hours of marathon meetings between those responsible for the Parliament, the Commission and the Council, the longest appointment in this type in the history of the Union, as its protagonists like to remember.

However, approving the law is only a “starting point”, stressed the other main rapporteur of the European Parliament, Dragos Tudorache. “Implementation is going to be key,” he indicated in reference to all the regulations that seek to regulate in some way the all-powerful technological platforms approved in this European mandate, which will conclude with the European elections in June. These include, in addition to the AI ​​law, the Digital Markets Act (DMA), an antitrust regulation that seeks to “put an end to the unfair practices” of giants of the digital economy such as Meta or Apple, and the Digital Services Act (DSA), which seeks to better protect consumers and their fundamental rights, combating, among others, misinformation or child pornography. According to Tudorache, once all these rules are “harmonized,” “greater certainty will be guaranteed to companies, but also to citizens,” when using new technologies.

You can follow The USA Print in Facebook and x or sign up here to receive our weekly newsletter.

_