Uploaded August 2026 | Updated September 2026, 1 hour ago
STW speaks with Anton Leicht, one of the most compelling voices on the geopolitics and economics of artificial intelligence. Anton has become a leading authority on how middle powers such as Australia can carve out a role in the global AI value chain and avoid being pushed to the periphery as the technology advances.
Anton is the author of the influential Substack Threading the Needle and a fellow with the Technology and International Affairs Program at the Carnegie Endowment for International Peace. He discusses Australia’s opportunity to build data centres to train and run AI, how that would help guarantee future access to advanced American models such as Claude and GPT, the prospects for sovereign models built here, and how Australia can sustain a cooperative relationship with the US built on mutual dependence.
The conversation then turns to AI and security: keeping models safe while avoiding a concentration of power in government hands, the responsibilities of the AI labs themselves, and compromise proposals such as third-party experts acting as safety watchdogs.
Anton closes with his views on the likely economic impacts of AI and the big questions around the future of work. He’s a frontier thinker on the political economy of AI, and is closely watched by policymakers, fellow experts and industry leaders alike.
STW speaks with Anton Leicht, one of the most compelling voices on the geopolitics and economics of artificial intelligence. Anton has become a leading authority on how middle powers such as Australia can carve out a role in the global AI value chain and avoid being pushed to the periphery as the technology advances.
Anton is the author of the influential Substack Threading the Needle and a fellow with the Technology and International Affairs Program at the Carnegie Endowment for International Peace. He discusses Australia’s opportunity to build data centres to train and run AI, how that would help guarantee future access to advanced American models such as Claude and GPT, the prospects for sovereign models built here, and how Australia can sustain a cooperative relationship with the US built on mutual dependence.
The conversation then turns to AI and security: keeping models safe while avoiding a concentration of power in government hands, the responsibilities of the AI labs themselves, and compromise proposals such as third-party experts acting as safety watchdogs.
Anton closes with his views on the likely economic impacts of AI and the big questions around the future of work. He’s a frontier thinker on the political economy of AI, and is closely watched by policymakers, fellow experts and industry leaders alike.
