Chapter 32: Quiz — Global AI Governance Frameworks

Instructions: Choose the best answer for each multiple-choice question. For short-answer questions, write 2–4 complete sentences. This quiz covers the main chapter content, both case studies, and the key takeaways.

Multiple Choice (Questions 1–15)

1. The OECD AI Principles, adopted in 2019, were significant primarily because they were: - A) The first legally binding international AI governance instrument - B) The first intergovernmental AI ethics agreement - C) The first AI governance instrument to cover environmental sustainability - D) The first AI governance instrument to address gender equality

2. The UNESCO Recommendation on the Ethics of AI (2021) was adopted by: - A) 38 OECD member states - B) G7 nations plus the EU - C) 193 UNESCO member states - D) 46 countries that formally adhered to the OECD AI Principles

3. The G7 Hiroshima AI Process was triggered primarily by: - A) China's deployment of national social credit scoring - B) The rapid deployment of GPT-4 and large language models - C) The EU AI Act's entry into force - D) Russia's use of AI in military operations in Ukraine

4. The "Brussels Effect" refers to: - A) The EU's influence over international standards bodies through its seat in Brussels - B) The mechanism by which EU regulation becomes de facto global regulation through market power - C) The EU's approach to blocking non-compliant AI products at its borders - D) The concentration of global AI governance institutions in Brussels

5. Which of the following best describes the legal status of the G7 Hiroshima AI Code of Conduct? - A) A binding treaty ratified by G7 member governments - B) A directive that must be implemented into national law by G7 members - C) A voluntary commitment without enforcement mechanisms - D) A soft law instrument with monitoring by an independent body

6. In the "three-bloc dynamic" of global AI governance, China's approach is best characterized as: - A) Regulation-first, focused on individual rights protection - B) Market-first, minimizing government intervention - C) State-control, using AI regulation to serve political stability and party priorities - D) Civil society-led, with governance driven by NGO input

7. The EU AI Act's extraterritorial scope means that it applies to: - A) Only AI companies headquartered within EU member states - B) AI companies that export their products to EU markets, regardless of where they are based - C) AI companies that receive EU government funding - D) Any company that employs EU citizens, wherever it is based

8. The Global Partnership on AI (GPAI) is hosted by: - A) The United Nations in Geneva - B) The World Economic Forum in Davos - C) The OECD in Paris - D) The European Commission in Brussels

9. Under the EU AI Act, AI systems used for real-time remote biometric identification in public spaces for law enforcement are categorized as: - A) Minimal risk, requiring no special governance - B) Limited risk, requiring transparency disclosures - C) High risk, requiring conformity assessment - D) Unacceptable risk, and are prohibited (with narrow exceptions)

10. The OECD AI Policy Observatory serves primarily what function? - A) Enforcement of OECD AI Principles against non-compliant member states - B) Monitoring and comparative analysis of AI policy across member and partner countries - C) Certification of AI systems meeting OECD AI Principles standards - D) Regulation of AI companies operating in OECD member countries

11. The AU AI Continental Strategy's concept of "data colonialism" refers to: - A) Africa's use of AI to monitor its citizens - B) EU data protection rules imposed on African companies without their input - C) The pattern of AI companies extracting data from African populations for training without adequate benefit-sharing - D) African governments using AI to colonize unregulated online spaces

12. Which of the following best describes the OECD's "race to the bottom" concern in AI governance? - A) Countries competing to develop AI capabilities faster than their rivals - B) Governments weakening AI governance standards to attract AI development activity - C) AI companies competing to deploy less capable but more widely accessible AI systems - D) The declining quality of AI ethics training programs globally

13. In Anu Bradford's analysis, the "Brussels Effect" operates primarily because: - A) The EU has enforcement authority over companies in any country - B) Companies find it cheaper to standardize to the strictest applicable standard - C) Other governments voluntarily adopt EU standards through diplomatic agreement - D) The EU market is so large that companies cannot afford to exit it

14. The UN High-Level Advisory Body on AI, convened in 2023, recommended: - A) A binding international treaty on AI governance modeled on the Nuclear Non-Proliferation Treaty - B) Expanding GPAI membership to include all UN member states - C) Establishing a new UN entity dedicated to AI governance and an international scientific panel on AI - D) Transferring AI governance authority from the ITU to UNESCO

15. The main limitation of civil society participation in global AI governance forums is best described as: - A) Civil society organizations generally oppose AI development and are therefore excluded - B) Resource constraints, language barriers, and venue accessibility limit meaningful participation - C) Government representatives are legally prohibited from taking civil society positions seriously - D) Civil society organizations lack the technical expertise to contribute to AI governance discussions

Short Answer (Questions 16–20)

16. Explain the concept of "regulatory arbitrage" in the context of AI governance. Give one specific example — real or hypothetical — of how a company might exploit regulatory arbitrage in AI, and explain what governance mechanism could address it.

17. The chapter describes the UNESCO Recommendation's universal adoption as both a significant achievement and a limited one. Explain this apparent contradiction: How can something be simultaneously historically significant and practically limited?

18. Describe the "three-bloc dynamic" in global AI governance and explain how each bloc attempts to shape global AI norms. Which mechanism do you think is most effective at shaping actual AI development practices, and why?

19. The chapter discusses the tension between soft law and hard law in AI governance. What are the two strongest arguments for relying primarily on soft law in international AI governance, and what are the two strongest arguments against it?

20. The AU AI Continental Strategy calls for African data sovereignty. What does this concept mean? What practical governance mechanisms could implement it? What obstacles would implementation face from major AI companies and foreign governments?

Answer Key: 1-B, 2-C, 3-B, 4-B, 5-C, 6-C, 7-B, 8-C, 9-D, 10-B, 11-C, 12-B, 13-B, 14-C, 15-B. Short answers should demonstrate understanding of the relevant concepts, engagement with the chapter's arguments, and application of concepts to specific examples.