Ep. 174: AI and policy, both foreign and domestic

In an episode recorded just before Christmas, Darren interviews Janet Egan, Senior Fellow and Deputy Director of the Technology and National Security Program at CNAS, about AI policy and its implications for Australia. Janet (who started her career in the Australian government) frames the current AI landscape as a two-horse race between the US and China, given vastly asymmetric investment levels. She introduces “compute policy” as a tractable governance lever, explaining that the physical infrastructure required for AI—specialised chips, data centres, and energy—offers regulatable chokepoints unlike easily transferable data or algorithms. The US strategy focuses on scale and removing barriers to advancement, while China, constrained by export controls on advanced semiconductors, pursues a diffusion-oriented approach emphasising open-source models and practical applications.

Turning to Australia's recently released National AI Plan, Janet offers a mixed assessment. She praises the establishment of an AI Safety Institute and the acknowledgment that data centres matter, while noting the plan avoided overly restrictive regulation that could stifle investment. However, she argues the plan misses a significant opportunity: positioning Australia as a compute hub for frontier AI training. Australia’s renewable energy potential, available land, and skilled trades workforce make it attractive for data centre buildout, but copyright restrictions on training data remain a key barrier.

Janet argues that unlike critical minerals, AI does not lend itself to hedging between Washington and Beijing given its inherently dual-use nature and emerging evidence of bias in Chinese models. She highlights the UAE and UK as instructive cases—the former for ambitious state-led mobilisation, the latter for sophisticated thinking about AI sovereignty structured around supply resilience, value capture, and strategic influence. For Australia, she argues, meaningful participation in the AI supply chain would provide strategic leverage and a seat at the table where consequential decisions are being made.

Australia in the World is written, hosted, and produced by Darren Lim, with research and editing this episode by Hannah Nelson and theme music composed by Rory Stenning.

Relevant links

Janet Egan (bio): https://www.cnas.org/people/janet-egan

Janet Egan, Spencer Michaels and Caleb Withers, “Prepared, Not Paralyzed:

Managing AI Risks to Drive American Leadership”, Center for New American Security, 20 Nov 2025: https://www.cnas.org/publications/reports/prepared-not-paralyzed

Janet Egan, “Global Compute and National Security: Strengthening American AI Leadership Through Proactive Partnerships”, Center for New American Security, 29 July 2025: https://www.cnas.org/publications/reports/global-compute-and-national-security

Lennart Heim, Markus Anderljung and Haydn Belfield, “To Govern AI, We Must Govern Compute”, Center for New American Security, 28 March 2024: https://www.cnas.org/publications/commentary/to-govern-ai-we-must-govern-compute

Emanuele Rossi, “Undersecretary Helberg explains Pax Silica and the Indo-Pacific AI play” Decode 39, 17 December 2025: https://decode39.com/12841/undersecretary-helberg-explains-pax-silica-and-the-indo-pacific-ai-play/

Stanford Institute for Human-Centered AI, “Artificial Intelligence Index Report 2025”, https://hai.stanford.edu/ai-index/2025-ai-index-report

Department of Industry, Science and Resources (Australia), National AI Plan, December 2025: https://www.industry.gov.au/publications/national-ai-plan

Helen Toner, “Rising Tide” (substack): https://helentoner.substack.com/

Lady Gaga, How Bad Do U Want Me (Official Audio): https://www.youtube.com/watch?v=Nd_M9A5xFlY