On Wednesday, March 13, 2024, the European Parliament endorsed the Artificial Intelligence Act, marking a pivotal moment in legislative history. This landmark piece of legislation aims to ensure the safety and adherence to fundamental rights in AI systems, while simultaneously fostering innovation within the European Union. The AI Act stands as the world’s inaugural comprehensive regulatory framework tailored specifically to govern artificial intelligence.
The Act classifies AI technologies into various risk levels, spanning from “unacceptable” to high, medium, and low hazards. Technologies falling under the “unacceptable” category will be outright banned, while high-risk AI systems will be subject to rigorous obligations. The regulation is slated to become effective at the culmination of the legislative term in May, subsequent to final checks and endorsement from the European Council. Implementation will commence in stages starting from 2025.
Notable provisions of the AI Act include:
- Prohibition of certain practices, enforceable six months post the law’s enactment.
- Establishment of codes of practice within nine months post the law’s enactment.
- Implementation of rules for general-purpose AI, including governance, within 12 months post the law’s enactment.
- Application of obligations for high-risk systems within 36 months post the law’s enactment.
Furthermore, the Act mandates transparency requirements for general-purpose AI systems, necessitates clear labeling of deepfakes, and establishes regulatory sandboxes to facilitate innovation, particularly supporting small and medium-sized enterprises (SMEs). The inception of the AI Act directly responds to proposals from citizens during the Conference on the Future of Europe (COFE), addressing the imperative for a safe and dependable society, human oversight of AI, and enhanced access to information for citizens, including those with disabilities.
The legislation will be fully enforceable 24 months post its enactment, encompassing specific provisions for general-purpose AI, biometric identification systems, bans on social scoring, and consumer rights to file complaints and receive explanations. The European Parliament’s approval of the AI Act signifies a momentous stride in AI regulation, establishing a model for other nations to emulate and striking a balance between innovation and safeguarding citizens’ rights.