On June 14, 2023, the European Parliament adopted negotiating position regarding the proposal for the Regulation of the European Parliament and Council laying down Harmonized Rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, the world’s first comprehensive legislation related to Artificial Intelligence (the “AI”).

According to the proposal, AI systems that can be used in different applications are analyzed and classified according to the risk they pose to users. The different risk levels will mean more or less regulation.

The priority is to make sure that AI systems used in the EU are safe, transparent, traceable, non-discriminatory and environmentally friendly.

The new rules establish obligations for AI providers and users depending on the level of risk from AI.

i)  Unacceptable risk

Unacceptable risk AI systems are systems considered a threat to people and will be banned. They include cognitive behavioral manipulation of people or specific vulnerable groups: for example, voice-activated toys that encourage dangerous behavior in children, real-time and remote biometric identification systems, such as facial recognition.

ii) High risk

Those are AI systems that negatively affect safety or fundamental rights, for example systems for biometric identification and categorization of natural persons, management and operation of critical infrastructure, law enforcement, assistance in legal interpretation and application of the law, and management of migration, asylum and border control.

iii) Limited risk

Limited risk AI systems will have to comply with minimal transparency requirements, which will allow users to make informed decisions. After having interacted with an application, the user is free to decide whether they want to continue using it. Users should be made aware of when they are interacting with AI. This includes AI systems that generate or manipulate image, audio or video content.

Generative AI will have to comply with transparency requirements, in particular it will be required to disclose that the content is AI-generated, to design the AI model to prevent it from generating illegal content, and to publish summaries of the copyrighted data used for training up these AI systems.

The aim is to reach an agreement on the proposal between the European Union member states by the end of the year 2023.