The European Union reached an agreement on Friday night to close the European artificial intelligence law. A pioneering rule in the world with which this technology is regulated for the first time based on the level of danger involved in its use. The negotiation was closed after three days of meetings.
“Historical! The European Union becomes the first continent to approve clear rules for the use of AI”, announced the Internal Market commissioner, Thierry Breton, on his account on the social network X. For his part, the president of the European Commission, Ursula von der Leyen, argued that the new law offers “a unique legal framework for the development of artificial intelligence that can be trusted”.
The point that most complicated the negotiations was that concerning biometric surveillance, such as facial recognition cameras in public spaces that use real-time artificial intelligence.
In a tug-of-war between the Eurochamber, which called for a total ban, and the states, which insisted on exceptions, a balance was finally reached. It will be prohibited in general, and will only be used in three cases: in the search for victims of kidnapping or human trafficking, to prevent terrorist attacks or when they are taking place; or to identify a person suspected of having committed crimes of terrorism, murder, rape or sex trafficking. For the negotiators of the Eurochamber it is “a success” (they admitted that it was not easy for the states) the fact that this system can only be used with prior judicial authorization.
States have insisted from the beginning on their use in cases of national security or to prevent cases of terrorism, but Parliament was worried that the fundamental rights of citizens could be violated.
On the other hand, a total ban on biometric categorization systems based on characteristics, such as ideology or sexual orientation, has been agreed upon. Also the recognition of emotions at work or in educational institutions. Likewise, systems that “score” citizens will not be allowed (known as social scoring), which consists of the fact that artificial intelligence tools give points to a person’s credibility or reputation according to their interactions on the internet or through a series of shots, a technology that is used massively in China.
Legislators also agreed on a list of systems considered high risk, and although they will not be prohibited, their use will be delimited in such a way that obligations will be applied to companies that want to enter the European market. Those with the highest risk will have safeguards and can only be used if it is guaranteed that they respect fundamental rights, such as the systems used in banking that assess the creditworthiness of a customer.
The European Commission proposed this law in 2021, when ChatGPT had not even been born and experts were not yet warning of its disruptive power. But the speed at which this technology advances made its regulation necessary, and this was a weighty reason for the negotiators of the Eurochamber and the governments, aware that not approving anything was not viable for either party.
Precisely, another of the issues that bothered me the most was the regulation of foundational models, generative artificial intelligence, such as ChatGPT, Bard or the recent Gemini, and which the Eurochamber asked to be included in this law. Against all odds, it was one of the agreements reached in the first meeting on Thursday night. Generative AI models such as ChatGPT will have to meet transparency requirements, such as systems respecting copyright and specifying whether content, such as photos or videos, is generated by AI.
Both institutions agreed that transparency requirements needed to be met, such as reporting when certain content has used this technology. It was important for countries to regulate it and put order, for example, in how these systems are used in the film industry and avoid a flood of lawsuits or strikes, such as the one recently called off in Hollywood in that the use of artificial intelligence has been one of the main battlehorses.
The negotiation table also discussed how they can affect technologies that create a big impact, such as deepfakes. Very realistic videos in which people appear talking and it is very difficult to distinguish whether they are real or created by artificial intelligence. Also altered images, as is the case of minors who were victims of manipulated images in September in Almendralejo. In both cases, it was agreed to include a brand that declares that the content is the result of AI.
The regulation establishes fines with a range that goes from 35 million euros or 7% of the global business volume of the companies, up to 7.5 million euros or 1.5% of the global business volume.
Due to the fact that it is a law that regulates a technology that is constantly evolving, and although it has the vocation to last, it will be allowed to introduce changes without having to re-reform the entire rule from top to bottom. “We have managed to maintain an extremely delicate balance: favoring innovation and strengthening artificial intelligence in Europe while fully respecting the fundamental rights of our citizens”, assured the Secretary of State for Digitization and Artificial Intelligence, Carme Artigas , in charge of leading the negotiations of the rotating Spanish presidency of the Council, after admitting that the negotiations were “tough and stressful”, but that “they were worth it”.
The puzzle to agree a pact was not only about closing technical aspects, although of great relevance, it was also necessary to find a balance between the benefits of artificial intelligence and the economic impact, and not lose competitiveness against giants like now China or the United States, which invest massively. While the European Union only represents 7% of annual investment, the United States and China together reach 80%. For this reason, both France and Germany insisted on the importance of this law not limiting competitiveness. Spain thus closes one of the key legislations of the European legislature and which marked one of its milestones for the presidency of the Council of the European Union.
After the agreement was reached, it must now be ratified by the states and the European Parliament (expectedly in February), and although it is expected to come into full force in 2026, before then it is expected that there have a gradual application. A new European office of artificial intelligence will be created at the time of its approval, which will coordinate the countries and will be advised by scientists and civil society. In six months, the ban on systems that will not be allowed will come into force, and the obligations of generative artificial intelligence models twelve months from the final approval.
This closes one of the most important laws of this legislature, marked as a milestone for the Commission, but also for Spain’s rotating presidency of the Council of the EU.
