The European Parliament calls for “an agile regulation” for artificial intelligence | Technology

The EU is losing the global race for technological leadership. It needs to increase the pace of investments in artificial intelligence (AI), which should reach around 20,000 million euros per year (now it is around 14,000) adding public and private spending. Efforts must also be redoubled in training, both in compulsory education and with digitization courses for adults. Only with an agile regulation, which establishes the necessary counterweights with the negative effects of technology, but without curtailing its capacity for innovation, will it be possible for European technology companies to emerge.

These are the main conclusions of the final report of the special committee of the European Parliament on Artificial Intelligence, approved today in Strasbourg and which establishes the position of the community legislative body on the subject. The document has no normative validity; it is rather a declaration of intent that comes at a time when the EU is fully developing its technological legislative framework. The European Commission is preparing the Artificial Intelligence Regulation that, among other things, applies restrictions to certain facial recognition systems or prohibits social credit systems. The Commission must take into account what the report that was approved today says.

The spirit of the document could be summed up in one sentence: it is important to regulate artificial intelligence, but a way must be found so that this does not prevent the development of a powerful European industry. A difficult balance to establish, but one that Europe wants to champion. “If we don’t set the standards, others will. We do not suggest regulating the technology, but its use”, highlighted Margarethe Vestoger, Vice President of the Commission and head of the European Digital Agenda, invited to the plenary session.

Several MEPs stressed during the debate that the essence of the EU is the protection of rights, and that technology should not be seen as a technical issue, but a political one. “When we talk about predictive policing, we have to know what impacts the presumption of innocence, and when we refer to biometric recognition technologies, we touch on the right to non-discrimination”, explained Juan Fernando López Aguilar, from the Socialist group.

“Our standards are based on values. But I think we must not lag behind, we must be agile”, said German MEP Axel Voss, from the European People’s Party, who led the preparation of the report. “Personally, I am not a fan of facial recognition, but leaving that technology completely aside will mean that others take the initiative and develop it in ways that we may not like.”

Other proposals included in the report are the creation of a European Technology Stock Exchange (the continental equivalent to the NASDAQ) that can attract capital for EU companies or the development of an interconnected data space.

Among the complaints made by the MEPs during the plenary session, two aspects stand out: that there is no call to strictly ban autonomous weapons (it is indicated that, ultimately, the final decision rests with a human being) or biometric identification systems, such as facial recognition (the authorities can resort to these systems in some cases in which citizen security is in danger).

A regulatory framework under construction

The Special Commission on Artificial Intelligence in the Digital Age (AIDA) began its work in September 2020. It was mandated to explore the impact of AI on the EU economy and its different sectors. The objective of the final report, which has no normative validity, is to establish a roadmap for AI within the EU until 2030.

These efforts are part of a wide framework of regulations, some of which are still under construction, which are shaping the community legal architecture in technological matters. To the General Data Protection Regulation (RGPD) of 2018, which establishes the basic principle that the data of European citizens cannot be collected or used without their consent, this year the Digital Markets Directive (DSA in its English acronym), which sets transparency standards for technology companies and fines for non-compliance, and the Digital Markets Directive (DMA), which establishes mechanisms so that bottlenecks are not generated in the market.

By next year, the Commission is expected to have the European Artificial Intelligence Regulation ready, the document that will regulate what use can be made of this technology. The approach adopted is that of risk: applications are classified according to the threats they pose to society and the development of people and the use of those classified as high risk is limited. In the draft directive, which can still be altered, artificial intelligence systems that act with medical data or facial recognition systems that act in public spaces, among others, fall into this category.

You can follow THE COUNTRY TECHNOLOGY in Facebook Y Twitter or sign up here to receive our weekly newsletter.