The AI Act

This post will be updated to contain the most recent information on the AI Act as well as links to other relevant articles. For more information on the process of how the AI Act came to be, take a look at Timeline and developments AI Act. 

Current text

The adopted text of the AI Act can be found here. The most recent official version of the text can be found here.

Definition of AI system

An AI system is a machine-based system designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.

High-risk AI systems

High-risk AI systems are classified based on Article 6 based on two main criteria: would the AI system be a safety component or covered by harmonization legislation or is it meant for one of the applications mentioned in Annex III. For more details on classification of high-risk systems, please have a look at our flowchart.

Timeline of implementation

March 2024: Final vote on the AI Act in the European Parliament.

March/April 2024: Publication in the Official Journal of the EU.

October 2024 (+6 months): Prohibited AI practices (Article 5) have to be removed from the market.

April 2025 (+12 months): GPAI models must be compliant.

April 2026 (+24 months): AI systems which fall under the AI Act but not under other harmonization legislation must be complaint.

April 2027 (+36 months): AI systems which fall under the AI Act and under harmonization legislation must be compliant.