Ресурсы

Since 2016, over thirty countries have passed laws explicitly mentioning AI, and in 2024, the discussion about AI bills in legislative bodies has increased globally. This policy brief aims to inform legislators about the different regulatory approaches to artificial intelligence (AI) being considered worldwide by legislative bodies.

The brief explains nine emerging regulatory approaches, each illustrated with specific cases worldwide. The order in which the nine AI regulatory approaches are presented is deliberately structured to guide readers from less interventionist, light-touch regulatory measures to more coercive, demanding approaches. These regulatory approaches are not mutually exclusive and AI bills often combine two or more approaches:

1. Principles-Based Approach

2. Standards-Based Approach

3. Agile and Experimentalist Approach

4. Facilitating and Enabling Approach

5. Adapting Existing Laws Approach

6. Access to Information and Transparency Mandates Approach

7. Risk-Based Approach

8. Rights-Based Approach

9. Liability Approach

The policy brief suggests parliamentarians how they can address three key questions before adopting AI regulations:

1. Why regulate? Determine whether regulation is needed to address public problems, fundamental and collective rights, or desirable futures.

2. When to regulate? Reach a consensus on why regulation is needed, map available regulatory instruments, compare them with other policy instruments, and assess the feasibility of adopting the former.

3. How to regulate? Identify a combination of AI regulatory approaches that are tailored to specific contexts.