Future Fund: AI world order defined by US-China split will hurt capital owners

A geostrategic standoff between the US and China over AI would be “extremely damaging for capital owners”, according to the Future Fund, Australia’s A$337 billion ($237 billion) sovereign wealth fund, which outlined four “secular scenarios” around AI and its investment implications.

In a glimpse into its scenario analysis process, the fund detailed potential world orders around AI which could see the technology either become a beneficial tool for transnational cooperation on issues like health or climate, or a great point of tension between global superpowers.

The fund wrote in its latest position paper that a “rapprochement” scenario will see a “unipolar” world where the US has a hegemonic advantage around AI. It will lead to a revitalisation of the US economy as it reaffirms its position as a global economic and policy leader, leading to high deployment scope around chip technology.

On the other end of the spectrum is a “divided world” scenario which the Future Fund believes could have devastating impacts for investors. It is defined by fierce geostrategic competition between the US and China, creating a highly bifurcated system across goods, people, capital and technology.

The fund believes this could lead to consequences including minimal productivity gains related to AI beyond defence, a “complete breakdown” of chip technology sharing, highly biased AI models as data is confined within blocs, and rampant cyberattacks and AI network disruptions.

“The process leading to a divided world could be extremely damaging for capital owners, particularly if abrupt,” the fund wrote in the paper. “Economies are heavily directed for geopolitical objectives, at the expense of economic and living standards objectives. High inflation is highly likely in the early years as capex is directed to defence.”

But in an interview with the local media, Future Fund chief executive Raphael Arndt said the world order right now is closest to what the fund called the “ruptures” scenario. The status is defined by an “uneven” distribution of power between states or blocs, which creates a high-tension environment where alliances shift based on self-interest.

The report warned that it could become a prelude to the “divided world” scenario with a higher chance of “strategic mistakes”.

“A broader range of inputs to production, output and capital flows are impeded. Stagflationary pressures from deglobalisation and efficiency loss emerge. Laissez-faire political systems [are] under pressure on growing need for state-directed national security and industrial policy,” the fund warned in the report.

The final case is the “resilience” scenario where there is a broad distribution of power between blocs and states. While it isn’t defined by complete deglobalisation, non-economic motives around government policies are increasing with the potential to greatly disrupt foreign direct investment.

In this scenario, middle powers may see the opportunity to boost defence capability via AI or improve security independent of US or China support.

However, one commonality across the four scenarios is the fund believes AI will increasingly be treated as a sovereign asset, and in extreme cases could lead to a complete nationalisation of critical AI assets as a way to shore up defence.

Portfolio implications

With the scenarios in mind, the Future Fund said the way to maximise the opportunity set in the AI thematic is maintaining access points across multiple asset classes. For example, the fund’s current exposures around AI assets include foundational models through venture capital, hyperscalers and multiple adopters through listed equity, data centres through infrastructure and quantum computing through venture capital.

“We have identified four broad layers that the current AI ecosystem can be defined into – hardware, infrastructure, platforms and applications,” the fund explained.

“AI, as an investment thematic, has a profound opportunity set attached with large upside potential.

“This comes from being a major driver of economic growth and thus investment returns over the long term through structural market beta tailwinds for AI-aligned markets, as well as the dispersion between AI leaders and laggards that creates opportunities for skilled investors to earn active returns – alpha.”

On the risk side, the Future Fund said it is most concerned about supply-side inflation pressure during the AI buildout phase, but as it becomes a more general-purpose technology it is likely to alleviate the pressure over time.

The fund is also keeping an eye on potential fallout from model convergence as machine learning models are increasingly trained on similar datasets and respond to related macro triggers. As the usage of AI models becomes pervasive in asset management and trade executions, the Future Fund said it “creates conditions where a single adverse signal could trigger synchronised deleveraging across multiple institutions before human intervention is possible”.

“To be successful in harvesting robust, repeatable returns through investing in AI, this will require a blend of foresight and adaptability, with winners and losers being difficult to discern across sectors, geographies and capital market landscapes,” the report said.

“Through thoughtful portfolio building and repositioning, the Future Fund has increased its exposure to the ‘engines’ of AI and broader technological revolution.”