AI Act - Key Deadlines and Compliance for Businesses

On 12 July 2024, the European Union (“EU”) has officially published the EU Artificial Intelligence Act (“AI Act”) within the Official Journal of the European Union marking a significant milestone in the regulation of artificial intelligence technology.

As upheld by Article 113 of the now published AI Act, 2 August 2024 will mark the entry into force of the AI Act, being the twentieth day after its publication.

It is to be noted that the published version of the AI Act contains minor differences from the latest draft. Such differences will be explored in an upcoming article.

The Timeline

The AI Act will not be completely enforced in its entirety and instead, a phased implementation will be followed to ensure a smooth roll-out which aims to promote comprehensive compliance and governance. The below figure illustrates the phased implementation:

By 2 February 2025, all prohibited AI practices must be withdrawn from the market;

By 2 May 2025, the AI Office Codes of Practice will be finalised;

The 2 August 2025 will mark the deadline for General Purpose AI models to comply with the Act, and for the establishment of the governance structure, including the AI Office, European Artificial Intelligence Board, and national market surveillance authorities;

On 2 February 2026, the European Commission will adopt the Implementing Act, detailing provisions for post-market monitoring plans.

By 2 August 2026, the entirety of the AI Act regulations will apply, including obligations for high-risk systems, and Member States must have at least one operational AI regulatory sandbox; and

Finally, by 2 August 2027, obligations for high-risk systems defined in Annex I will come into effect, further ensuring the Act’s comprehensive enforcement.

With the imminent enforcement of the AI Act, in scope entities operating within the EU will be required to align their AI practices with the new regulations. Compliance to avoid the significant imposed penalties will be of paramount importance, especially those entities producing AI products falling under the obligations for high-risk systems as defined in Annex III of the AI Act,

The AI Act is expected to be the lynchpin of AI regulation worldwide. By ensuring greater accountability and imposing the responsible and transparent AI development, will foster both a trustworthy and an ethical AI ecosystem the EU and beyond.

Author: J.J. Galea

Do you require assistance complying with the imminent enforcement of the AI Act? GTG is here to help! Do not hesitate to contact Dr Ian Gauci for further clarification or assistance.