Exercises — Chapter 19: Global Perspectives on AI

Part A: Conceptual Questions

A1. ⭐ Define techno-nationalism in your own words. Give one historical example (before AI) and one current example.

A2. ⭐ What is the Brussels effect? Explain it using a non-AI example (like GDPR) before applying it to AI governance.

A3. ⭐⭐ Explain the difference between the U.S., Chinese, and European approaches to AI development in 2–3 sentences each. What values or priorities does each approach prioritize?

A4. ⭐⭐ What is digital sovereignty? Why might a country want to ensure that data generated by its citizens is stored within its borders?

A5. ⭐⭐ The chapter introduces the concept of the "compute divide." In your own words, explain what this means and why it matters for global AI equity.

A6. ⭐⭐⭐ Evaluate the claim that calling U.S.-China AI dynamics a "race" is misleading. What does the race metaphor get right? What does it get wrong?

A7. ⭐⭐⭐ The concept of "data colonialism" draws an analogy between historical colonialism and modern data extraction. Identify two ways the analogy is illuminating and two ways it might be misleading or overstated.

A8. ⭐⭐ Why does the chapter argue that AI literacy is especially important when considering global AI governance? What can an AI-literate citizen contribute that a technically illiterate one cannot?

Part B: Applied Analysis

B1. ⭐⭐ Choose one of the four AI governance approaches from the Debate Framework in section 19.6 (binding treaty, regulatory competition, industry self-regulation, multi-stakeholder). Identify a real-world precedent from another policy domain (environment, trade, nuclear weapons, etc.) where a similar approach has been tried. What can we learn from that precedent?

B2. ⭐⭐ Apply the concept of algorithmic monoculture to social media content moderation. If most major platforms use similar AI-powered moderation tools developed in the same country, what are the risks? What might a more diverse ecosystem look like?

B3. ⭐⭐⭐ A technology company headquartered in the United States wants to deploy a facial recognition system in three countries: Germany, Nigeria, and Singapore. Using what you have learned in this chapter, describe the different regulatory environments, cultural concerns, and practical challenges the company would face in each location.

B4. ⭐⭐ A government in the Global South is considering purchasing a "smart city" package from a Chinese technology company. The package includes traffic management, surveillance cameras with facial recognition, and predictive policing tools. Identify three questions the government should ask before making this purchase.

B5. ⭐⭐⭐ Research one AI initiative from the Global South (e.g., Masakhane NLP, India's Aadhaar system, Kenya's M-Pesa, Brazil's AI governance framework). In 200–300 words, describe what it does, who it serves, and what it reveals about AI innovation outside the U.S.-China axis.

Part C: Research Design & Critical Thinking

C1. ⭐⭐⭐ Design a study to test whether AI content moderation systems perform differently across languages. What languages would you test? What metrics would you use to measure performance? How would you control for the type of content being moderated?

C2. ⭐⭐⭐ The chapter presents the AI governance gap — the mismatch between AI's global reach and national-level regulation. Propose a governance mechanism that could help close this gap. Be specific about: (a) who participates, (b) how decisions are made, (c) how rules are enforced, and (d) how the mechanism adapts as technology evolves.

C3. ⭐⭐⭐⭐ Some scholars argue that the Global South should focus on "appropriate AI" — AI solutions designed for local contexts with local resources — rather than trying to compete with the U.S. and China on frontier models. Evaluate this argument. What are its strengths? What are the risks of this approach?

C4. ⭐⭐⭐ The EU AI Act bans social scoring systems but allows many other forms of AI-driven assessment (credit scoring, hiring algorithms, etc.). Is this distinction principled or arbitrary? Make an argument for where the line should be drawn, and justify it using concepts from this chapter and Chapter 13.

Part D: Synthesis

D1. ⭐⭐⭐ Write a 300-word op-ed arguing either for or against data localization laws. Consider the perspectives of at least three stakeholders: a government, a technology company, and ordinary citizens.

D2. ⭐⭐⭐⭐ The chapter discusses three AI governance strategies: innovation-first (U.S.), regulation-first (EU), and state-directed (China). Design a fourth strategy that a mid-sized democracy in the Global South might adopt. What would its priorities be? How would it balance innovation, regulation, and sovereignty?

D3. ⭐⭐⭐ Return to the ContentGuard example from section 19.4. If you were advising a social media company on making its content moderation system more equitable across global contexts, what three specific recommendations would you make? For each recommendation, identify one potential obstacle to implementation.

D4. ⭐⭐⭐⭐ The chapter quotes Melvin Kranzberg: "Technology is neither good nor bad; nor is it neutral." Apply this principle to the global dynamics of AI development and deployment. In 400–500 words, explain what Kranzberg's insight means in the context of AI geopolitics and offer your own position on whether AI can be made more "neutral" through better governance.

Part M: Mixed Practice (Spaced Review)

These questions revisit concepts from earlier chapters in the context of global AI dynamics.

M1. ⭐⭐ (Ch.9 connection) In Chapter 9, we discussed how bias enters AI systems through training data. How might this problem be amplified when an AI system trained primarily on data from one country is deployed in another? Give a concrete example.

M2. ⭐⭐ (Ch.12 connection) In Chapter 12, we explored privacy and surveillance. How do global dynamics complicate privacy protections? Consider a person in a country with strong privacy laws whose data is processed by a company in a country with weak ones.

M3. ⭐⭐ (Ch.13 connection) In Chapter 13, we studied national AI governance frameworks. What is the biggest limitation of governing AI at the national level when AI systems operate globally?

M4. ⭐⭐⭐ (Ch.17 connection) Chapter 17 examined AI and justice. How do the global inequities discussed in this chapter relate to the justice frameworks from Chapter 17? Can an AI system be "just" domestically but "unjust" globally?