Alternative reality

China has advanced a framework that empowers countries to explore and pursue their own vision of AI development

For the best part of the last three or four decades, US bigtech has in effect been global bigtech. US companies have stamped their leadership in hardware, software and applications delivering and controlling operating systems, platforms and services that have by and large fulfilled the role of a global technological ecology. The United States-based research labs have cornered the market for global talent, enabling the perspectives of US companies to dominate the mechanisms by which global technology standards have been set.

The hallmark of US bigtech has been a business model that incorporates proprietary technologies with substantial linkages to government finance. Silicon Valley wouldn't be what it is today without deep and lasting connections to the US military-industrial complex. Pentagon contracting laid the groundwork for Silicon Valley's renaissance in the 1990s, launching the current generation of dominant technology players.

US dominance of the technological landscape has enabled the US state to exercise considerable influence over the institutions that enable global commerce, enabling the US state to weaponize foundational digital infrastructure. The expansion of US bigtech over the past three decades enabled it to be turned into "tools of domination "by Washington as documented by US researchers Henry Farrell and Abraham Newman in their book Underground Empire.

And now, as we head into an era that's likely to be dominated by the development and application of artificial intelligence, a similar pattern of institutional configuration is evident in the US.

While much of the Pentagon's annual appropriations still goes to conventional weapons systems, and the usual grab-bag of defence sector contractors, a new political economy is emerging that entangles the threads of bigtech, venture capital and private equity. According to a 2024 study by Roberto Gonzalez for the Costs of War Project, the centrality of data has focused the minds of Pentagon's planners. For him, this "coupled with years of 'AI hype', generated by tech leaders, venture capitalists and business reporters… played a crucial role in sparking the interest of military leaders who have come to view Silicon Valley's newest innovations as indispensable warfighting tools".

It's no surprise that the "Princes "of Silicon Valley occupied pride of place at Donald Trump's inauguration.

The development of AI in the US has largely been cloistered within a proprietary environment. Venture capital and private equity finance promotes a technology business model that aims for sector dominance; "category killer" is the catchphrase, which encapsulates the idea that technology platform monetization is premised on the ability to eradicate competition. US bigtech is akin to the modern-day rentier, described separately as a form of technofeudalism by Yanis Varoufakis and Cédric Durand.

The US state, under the new Trump administration, is doubling down on this foundational political economic model. It sees US AI dominance as central to its own ability to maintain global primacy, and is willing to exploit its control over key elements of the total technology stack to enforce its own priorities and imperatives. Control over key semiconductor design technologies, critical algorithms and applications through corporations such as OpenAI (the developer of ChatGPT), Alphabet, Meta and Elon Musk's various enterprises, is seen to be pivotal to dictating who can access what technologies and under what conditions.

The US speaks of a need to enable AI development to flourish in a low-regulation environment, believing that concerns about AI and "safety" would be impediments to its development. Minimal regulatory constraints in this case is a proxy for US bigtech being relieved of the need to comply with regulatory requirements demanded by non-US jurisdictions. Again, the modus operandi is to enforce an environment in which US technology becomes de facto global technology.

The US speaks of a world of AI free of ideological bias all the while promoting an AI vision that is fully invested in the biases of US exceptionalism and primacy. This has been the story of US cultural and political hegemony through the ages. A rhetoric of ideological non-bias is pitted against a reality of the opposite, and this is no more evident than the very US notion of what constitutes an appropriate limitation of jurisdictional authority.

While new Secretary of State Marco Rubio speaks of unipolarity as an unnatural state, and the reality of multipolarity, US bigtech and associated government policies and strategies remain trained on a model that seeks to perpetuate US technological dominance. This dominance underpins the US' ability to exert control over the affairs of other countries, and expropriate resources and wealth from others under fear of being punished.

Modern day sanctions and associated punitive efforts are made possible because of US domination of the technological systems that enable global trade and finance. The US has successfully prosecuted regime change operations across the globe, enabled by its control over the information production and distribution networks, as recently revealed in the US Agency for International Development imbroglio. A proprietary AI, aligned to the interests of the US state, untrammeled by internationally agreed governance norms, is a means of perpetuating these capabilities under new conditions.

If the US' intentions are insufficiently clear, or are too opaque behind the glossy rhetoric of innovation, freedom and the absence of censorship, its fundamental strategic intent is laid bare by the fact that the US together with the United Kingdom refused to sign the Statement on Inclusive and Sustainable AI for People and the Planet.

Signed by 60 others, including France, China and India, the declaration pledged an "open", "inclusive" and "ethical" approach to AI's development. The statement aims to promote AI accessibility in a "transparent", "safe" and "secure and trustworthy" way, "making AI sustainable for people and the planet".

When one brings the threads together, we can see a US AI and technology agenda that speaks to unrestrained US ambitions for global primacy, and which subordinates accessibility under US capacity to control the entire technology stack. The US approach is not interested in accessibility, except on its own terms.

National sovereign development including in the area of AI requires a different framework. China has advanced a framework that aims to promote a genuinely open approach to AI development and adoption, which acknowledges and respects national sovereignty. This is a genuine contribution to global discussions about multilateral governance of critical public infrastructure and capabilities. Rather than seek to impose another ideological vision on other countries, China's proposal lays the foundation for countries around the world to be empowered to explore and pursue their own vision of AI development within a framework that acknowledges the need to do so in a responsible manner.

Nations now face distinct visions for the governance of the development and use of AI. On the one hand, there is the model of the techno-rentier, anchored by proprietary technology models, aligned with the interests of a single nation state. This is US exclusivity in practice. On the other hand, we have a multilateral governance framework that balances accessibility, and national sovereignty empowerment with a need for shared responsibility to develop the technology carefully. This is a multipolar inclusive approach.

The launch of DeepSeek as an open source platform, coupled with the fact that the model can run on non-US developed and manufactured chips, proves that the world now has real options.

The author is an adjunct professor at Queensland University of Technology and a senior fellow at the Taihe Institute. The author contributed this article to China Watch, a think tank powered by China Daily. The views do not necessarily reflect those of China Daily.

Contact the editor at editor@chinawatch.cn.