Week 2
IN THIS LESSON
-
Read only the following sections:
Abstract
Introduction (p. 1-7)
Overview of different regulator approaches (p. 20-27)
Proposing a regulatory approach for frontier AI (p. 49-54)
Optional: Depending on interest, you might want to skim certain parts from ‘Choosing a regulatory approach for frontier AI’ (p. 35-49).
Governments are beginning to regulate frontier AI systems by requiring developers to implement safety measures, with approaches ranging from high-level principles to specific rules. This report argues that 'principles-based' and 'rules-based' approaches should be viewed as a spectrum rather than a binary choice.
-
This RAND report provides a comprehensive primer on the European Union's AI Act and talks about its implications for US-based companies.
-
Watch this short explainer video by the Center for New American Security
The Trump administration recently released its AI Action Plan which outlines a comprehensive strategy to maintain US global AI dominance through three pillars: (1) accelerating innovation by removing regulatory barriers and promoting American values in AI systems, (2) building AI infrastructure and (3) leading international AI diplomacy while restricting adversaries' access to advanced AI technology.
-
Skim through the text of the Action Plan and read experts’ in-line comments on its most notable statements
-
China's AI governance evolution ended up creating a framework built on three pillars: content moderation, data protection, and algorithmic governance.
A primer on China’s Global Initiative for AI Governance, in particular pointing to how this complements similar initiatives from other leaders in global AI governance.
-
The first article explains the EU AI Act's Code of Practice for general-purpose AI (GPAI) model providers, which serves as a non-binding but crucial interim compliance framework bridging the gap between GPAI obligations taking effect in August 2025 and formal European standards arriving in 2027 (or later).
-
The author argues that AI middle powers (advanced economies excluding the US and China) should avoid trying to compete directly in AI development and instead focus on leveraging bottlenecks to maintain strategic relevance.
