European Commission: Proposal for an Artificial Intelligence Act   

IRIS 2021-6:1/25

Ronan Ó Fathaigh

Institute for Information Law (IViR)

On 21 April 2021, the European Commission published its landmark and much-anticipated Proposal for a Regulation laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). This follows the publication of the European Strategy on Artificial Intelligence in 2018, and the publication of the Guidelines for Trustworthy AI in 2019 by the High-Level Expert Group on Artificial Intelligence, established by the European Commission (see IRIS 2019-7/3). The 108-page Proposal for an Artificial Intelligence Act runs to 85 Articles and, as the Commission states, is based upon a risk-based approach, by prohibiting certain AI systems with unacceptable risks, while subjecting high-risk AI systems to strict obligations. 

At the outset, Article 3 of the Artificial Intelligence Act defines an AI system as software that is developed with one or more of certain techniques and approaches (listed in Annex I), such as “machine learning”, and can, for a given set of human-defined objectives, “generate outputs such as content, predictions, recommendations, or decisions influencing the environments they interact with”. Importantly, Title II of the Act then sets out a number of AI practices that are prohibited. For example, AI systems that “materially distort a person’s behaviour” in a manner that causes or is likely to cause that person or another person physical or psychological harm are prohibited under Articles 5(a) and 5(b). Further, AI systems that allow “social scoring” by governments (i.e., classification of the trustworthiness of individuals over a certain period of time based on their social behaviour) are also prohibited under Article 5(c). Notably, Article 5(d) prohibits the use of real-time remote biometric identification systems in publicly accessible spaces for the purpose of law enforcement, except in certain circumstances. These include the detection, localisation, identification or prosecution of a perpetrator or suspect of certain criminal offences, and the prevention of a specific, substantial and imminent threat to the life or physical safety of natural persons, or a terrorist attack. 

The Artificial Intelligence Act then addresses AI systems that are “high-risk” in Title III. These types of AI systems are contained in Annex 3 to the Act, and include AI systems for (a) education training; (b) employment, workers management and access to self-employment; (c) access to and enjoyment of essential private services and public services and benefit; (d) law enforcement; (e) assisting judicial authorities in researching and interpreting facts and the law and in applying the law to a concrete set of facts; (f) migration, asylum and border control management. Crucially, the Artificial Intelligence Act sets out a number of rules which apply to these high-risk AI systems, including rules on data and data governance, documentation and recording keeping, transparency and provision of information to users, and human oversight.

Of particular note, Article 52 lays down rules on deep fake technology, and provides that AI systems that generate or manipulate image, audio or video content that appreciably resembles existing persons, objects, places or other entities or events and would falsely appear to a person to be authentic or truthful (“deep fake”), must disclose that the content “has been artificially generated or manipulated”. However, this rule does not apply where use of the AI system is necessary for the exercise of the right to freedom of expression and the right to freedom of the arts and sciences, guaranteed in the EU Charter of Fundamental Rights, and “subject to appropriate safeguards for the rights and freedoms of third parties”. 

Finally, Title VI sets up a governance system at EU and member state level. First, Article 56 establishes a European Artificial Intelligence Board, which will provide advice and assistance to the European Commission on the application of Artificial Intelligence. Further, under Article 59, member states will be required to designate national competent authorities  for the purpose of ensuring the application and implementation of the Artificial Intelligence Act. 

The European Parliament and the Member States will now consider the Commission's proposal in the ordinary legislative procedure. If adopted, the Regulation will be directly applicable across the EU.