The EU AI Act: A comprehensive deep dive

The AI Act is expected to be published in the Official Journal on 12 July 2024, with its entry into force 20 days thereafter. This will mark the beginning of a phased implementation process to put the various rules and obligations of the AI Act into practice. For businesses and organisations, this means there is now a critical window to prepare for compliance. Read more in our EU Artificial Intelligence Act - Deep Dive,  please download the full PDF at the top of this page.

Putting the AI Act into practice

The AI Act establishes a framework for regulating the provision, deployment and usage of AI within the EU. It sets standardized processes for the market entry and operational activation of single-purpose AI (SPAI) systems, ensuring a cohesive approach across EU Member States. As a product safety regulation, the AI Act adopts a risk-based approach, categorizing AI systems based on their use cases and the level of risk they pose to users.

This categorization leads to specific compliance requirements, including bans on certain AI applications deemed unethical or harmful, and detailed requirements for high-risk AI applications to manage potential threats effectively. Additionally, it provides transparency guidelines for AI technologies designated to have limited risk.

AI ethics are at the heart of the AI Act, which aims to remain adaptable to future iterations of AI technologies. The public use of general-purpose AI (GPAI) technology necessitated differentiation between single-purpose AI and general-purpose AI. Consequently, the AI Act also regulates the market entry for general-purpose AI models, establishing comprehensive rules for market oversight, governance, and enforcement to uphold integrity and public trust in AI innovations.

Read more in our EU Artificial Intelligence Act - Deep Dive , please download the full PDF at the top of this page.

Navigating the future

Given its somewhat abstract nature, the AI Act includes areas yet to be fully defined. These areas are expected to be elaborated through delegated and implementing acts, guidelines by EU institutions, and harmonized standards developed by European Standardization Organizations. Businesses and organisationscan anticipate receiving more detailed guidance in the near future. We will of course keep you updated on these developments.

At Deloitte we believe in approaching the AI Act in a multi-disciplinary manner, bringing together legal, risk, ethics and technical expertise to support a successful, comprehensive, effective and efficient analysis and implementation.