The Artificial Intelligence Act (AIA) defines high-risk AI systems as AI systems that “pose a significant risk to safety or fundamental rights of natural persons, particularly in the following areas:
Safety of individuals: AI systems that are used to control or manage critical infrastructure, such as healthcare systems, transport systems, and energy systems;
Privacy and fundamental rights: AI systems that are used for automated decision-making, such as those that are used in recruitment, loan applications, and criminal justice;
Social biases and discrimination: AI systems that perpetuate or exacerbate social biases and discrimination;
Manipulation and disinformation: AI systems that are used to spread disinformation or manipulate people's behavior;
Child rights: AI systems that are used for targeted advertising to children or that collect personal data from children without parental consent.