AI and Digital Governance: Six Questions (and Answers) to Navigate the Complexities

Why It Matters

How governments define the object, timing and scope of AI rules will determine whether regulation mitigates harms (bias, privacy, safety) without stifling innovation, and will shape national competitiveness and international alignment in a rapidly advancing technology landscape.

Summary

Matthew presents a practical 5W1H framework (What, Why, Who, When, Where, How) to systematize AI and digital governance, arguing policymakers must decide whether to regulate data, models, or specific applications and to tailor rules accordingly. He reviews recent global developments: the EU AI Act, U.S. federal churn and state-level laws like California’s SB53 and New York’s RAISE Act, China’s mandatory AI content labeling, the Paris summit, and new UN and industry initiatives. The talk highlights trade-offs between safety, innovation and geopolitical competition—pointing to frontier models, agentic systems, and data provenance as central regulatory concerns. Matthew emphasizes that coherent answers across the six questions are needed to avoid piecemeal, inconsistent, or ineffective regulation.

Original Description

What is at stake in regulating AI in digital governance? As AI and digital technologies advance rapidly, governance frameworks struggle to keep pace with emerging applications and risks.

In this talk, Professor Matthew Liao, New York University, moderated by Professor Janina Dill, Blavatnik School of Government, will ask six questions to analyse governance frameworks:

1. What should be regulated (data, algorithms, sectors, or risk levels)?

2. Why regulate (ethics, legal compliance, market failures, or national interests)?

3. Who should regulate (industry, government, or public stakeholders)?

4. When should regulation occur?

5. Where should it take place (local, national, or international levels)?

6. How should regulation be enacted (hard versus soft regulation)?

Professor Liao will compare the European Union's AI Act to US regulation, revealing key differences.

Blavatnik School of Government,

University of Oxford

http://www.bsg.ox.ac.uk/