The EU AI Act Takes Effect in 2024

The EU AI Act is set to take effect in 2024 and adopts a risk-based approach, categorizing AI systems based on their potential impact and imposing proportionate regulatory measures.

Unacceptable and High-Risk AI Uses

The EU AI Act specifies that unacceptable-risk uses of AI are prohibited, including subliminal manipulation, exploitation of vulnerable groups, social scoring, and some real-time use of biometrics by law enforcement.

High-risk AI uses under the EU AI Act include products already regulated by the EU such as machinery, toys, medical devices, and transportation vehicles; safety components of critical infrastructure; and products or services impacting education, employment, public benefits, credit, law enforcement, border control, or judicial or democratic processes.

Fundamental Rights Impact Assessment Requirements

Article 27 of the EU AI Act introduces a Fundamental Rights Rights Impact Assessment (FRIA) requirement before deploying high-risk AI systems, which must be conducted by deployers that are public bodies or private operators providing public services.

Enforcement and Penalties

Penalties for violation of the EU AI Act range upwards to as much as 6% of global annual revenues.

Origins of the Legislation

The EU’s Artificial Intelligence Act, first introduced in 2021, aimed to create a harmonized framework for AI regulation, focusing on high-risk applications and promoting trustworthy AI practices.

Source: Foundations for Risk Assessment of AI in Protecting Fundamental Rights