The keys to the EU’s new Artificial Intelligence law

Did you know that the EU will be the first to establish clear rules for the use of Artificial Intelligence? Here’s what they will consist of.

Last December, the Council and the European Parliament reached a provisional agreement on what is the first Artificial Intelligence (AI) Regulation. This law stems from the European Commission’s 2021 proposal to create the EU’s first regulatory framework for AI.

The law has two clear objectives:

• To ensure that Artificial Intelligence systems used in the EU and introduced in the European market are safe and respect citizens’ rights.
• To stimulate investment and innovation in the field of AI in Europe. The agreement stipulates that the Artificial Intelligence Regulation should be implemented two years after its entry into force.

Importantly, this is the first legislative proposal on Artificial Intelligence in the world, so it can establish itself as a global benchmark for regulating AI in other jurisdictions. The law will regulate AI based on its risks. The higher the risk, the stricter the rules will be.

What is the first thing to know about this regulation?

Firstly, the interim agreement has clarified the OECD’s proposed approach as the definition of an AI system, with the aim of establishing clear criteria to distinguish AI systems from other, simpler software.

Furthermore, it is clarified that the Regulation will not apply to areas outside the scope of EU law and will not affect Member States’ national security competences. It also does not apply to systems used exclusively for military or defence purposes; nor does it apply to systems used for research and innovation purposes.

High-risk AI systems and prohibited practices

The agreement establishes a set of horizontal protection criteria that will determine when AI systems can cause harm to society. Thus, those with high risks will be subject to a number of requirements and obligations in order to gain access to the EU market.

In the case of systems with risks deemed unacceptable, their use will be banned altogether. These include, for example, cognitive behavioural manipulation or the indiscriminate tracking of facial images taken from the internet, among others.

Protection of fundamental rights

The regulation provides for the assessment of the impact of any new AI system on fundamental rights before it is introduced on the market. In addition, in the case of functional models, understood as large-scale systems that can perform different tasks (generation of video, text, images, etc.), they will have to comply with more specific transparency obligations.

Furthermore, it emphasises the obligation for users of an emotion recognition system to inform natural persons to whom such a system is exposed.

The European Commission’s IA Office

In order to ensure compliance with the new regulations, the Commission’s AI Office is to be set up. It will be responsible for monitoring Member States’ implementation of the new regulations, overseeing state-of-the-art AI models and helping to promote standards and testing practices.

More information: Representation of the European Commission in Spain