Summary
Regulatory agencies generally do not issue regulations or guidance on specific technologies, but instead address the importance of effective risk management, governance, and controls regarding the use of technology
There's a lot happening in the regulatory landscape, but clear themes are emerging. By analyzing current and proposed regulations, organizations can prepare to address these key areas:
- Clearly define what constitutes an AI system.
- Focus on managing high-risk impacts, particularly through governance and risk assessments.
- Differentiate between developers and deployers of AI systems.
- Ensure compliance with notice and transparency requirements.
- Protect consumer rights, such as allowing opt-outs from certain types of decisions.
Regulations need:
- Protections for research.
- Framework that takes priority when there are conflicts with other laws.
- Prohibit on unacceptable risk.
| Region | Approach | Features |
|---|---|---|
| US | Piecemeal regulation | Different agencies regulate AI in different ways, focus on specific sectors or applications.
Expect that states will lead and some of these regulations might conflict. |
| Japan | Voluntary guidelines | Focuses on promoting responsible development of AI, public-private partnership |
| EU | Comprehensive regulation | Focuses on high-risk AI applications, requires companies to take steps to mitigate risks |