Countries Seeking To Develop AI Capabilities Need To Navigate Trusted Partner Networks, Says Expert
By Siti Radziah Hamzah
TASHKENT, June 18 (Bernama) -- Countries seeking to develop artificial intelligence (AI) capabilities will need to navigate new alliances, controls and trusted partner networks as the technology increasingly becomes a geopolitical issue, said an expert today.
Tony Blair Institute for Global Change chief AI and innovation officer Benedict Macon-Cooney said recent developments, including United States (US) export controls on advanced AI models, underscored how access to AI technologies is increasingly shaped by national interests and geopolitical considerations.
"You have to think of AI as a geopolitical system, not just a technology today.
"Now what I think that looks like in terms of what it means to govern is that if you are a political leader or working at the forefront of this space today, you have to navigate new sets of alliances, new sets of controls and new sets of trusted partner networks," said Macon-Cooney.
He said this during a session titled "AI, Digital Governance and Innovation: The Next Investment Frontier" at the Tashkent International Investment Forum (TIIF) 2026.
Macon-Cooney said the global AI race is currently dominated by the US and China, with only a handful of countries possessing significant capabilities.
"I think it is clear at the top that the US and China are by far ahead in this AI race. You have the leading frontier labs such as Anthropic and OpenAI, and of course Google with its Gemini model, which are really at the frontier right now,” he added.
Macon-Cooney said China, supported by companies such as DeepSeek and other open-source AI developers, is estimated to be six to nine months behind the leading AI models from the US.
He noted that while countries such as the United Kingdom, France, Germany and Japan possess AI capabilities, they remain behind the two leading powers.
As a result, governments and policymakers would increasingly need to navigate new alliances, controls and trusted partner networks in accessing critical AI technologies.
Macon-Cooney cautioned that some countries risk misunderstanding AI sovereignty by assuming it requires building their own frontier models.
"I think for many the question then becomes one of sovereignty, but I think there is a very, very deep risk in how some nations are treating sovereignty today," he said.
Macon-Cooney likened AI development to establishing a national airline, where countries purchase aircraft from established manufacturers rather than building them from scratch.
"I think there is a profound mistake that some countries are making thinking that, there are some controls and questions about trusting some of these nations today, but you cannot simply start by building a frontier model, which might cost a trillion dollars plus to train from scratch, and you are going to have to think about how you ally with the most powerful nations on these capabilities," he added.
Speaking to Bernama later, Macon-Cooney said the same principle applies to middle-power economies seeking to strengthen their AI capabilities.
"The question always then, as you say, for middle powers is how do we compete when trillions of dollars of capital is being put into the frontier models? And I think the question of sovereignty then becomes very important in thinking about where you are going to have comparative advantage in the market.
“I think for most nations, building a frontier model is going to be out of the question," he opined.
Instead, he said countries should focus on strengthening supporting infrastructure, including data centres, computing capacity and energy resources, while leveraging AI technologies developed by frontier laboratories to improve productivity and competitiveness.
Macon-Cooney said countries could derive significant benefits by applying AI across sectors such as healthcare, life sciences, education and tax administration.
He added that while investments in data centres and digital infrastructure remain important, the long-term value of AI would ultimately depend on how effectively countries deploy the technology throughout the economy.
"I think there is still trillions of dollars being committed to capital infrastructure and building data centres and some of that wider infrastructure. And I think there is still a long way to go on that build-out, including underlying bottlenecks such as energy, chips and memory," said Macon-Cooney.
Beyond digital services, he said AI would increasingly be deployed in manufacturing, advanced robotics, drug development and biological research, creating opportunities for countries that do not possess frontier AI labs but are able to build industrial capacity around the technology.
"But other nations that do not have frontier labs will begin to think about how do we build that industrial capacity," he said.
-- BERNAMA