Tech sovereignty - you can have AI in any color you like, so long as it’s red, white and blue

The number-one factor that will define whether the US or China wins this race is whose technology is most broadly adopted in the rest of the world.

Those were the words of Microsoft General Counsel Brad Smith at a US Senate hearing last May. Since then the topic of tech sovereignty has gained more and more currency, not least in relation to AI, with the debate largely coming down to the US and/vs - delete as applicable according to your worldview - Europe and the growing divide between the two.

At the World Economic Forum gathering in Davos last month, for example, sovereignty was never far from the conversation.

The transatlantic gulf has not been helped by inflammatory commentary on both sides of the Pond. Sticking with the US for the moment, statements like this from Vice-President JD Vance are hardly likely to calm nerves:

The Europeans, they’re so friendly in private and they’re willing to make a lot of accommodations and then publicly they attack us and they say, ‘We’re not going to work with the Americans. We’re not going to do anything with the Americans’. I’m sorry - it’s all bogus.’

Of course, the counter-argument from the Europeans - which we’ll look at more closely in a follow-on article - would be to cite such commentary as a good reason to look to break the umbilical cord with the US and put European tech sovereignty into practice, something, as we shall see, much easier said from the comfort of a Davos conference room than actually actioned.

The fear at present is that when it comes to matters such as regulation and standards around emerging technology, the US stance is akin to that of Henry Ford and his maxim that car buyers could have any color vehicle so long as it was black. Update that to, you can any kind of tech sovereignty you want so long as it’s red, white, and blue!

Charm 

But the US has recently been on something of a charm offensive to convince the rest of the world, except China, that this isn’t the case. This week, Michael Kratsios, Chief Technology Officer of the United States, a man who advises President Trump on a tech policy issues, gave an interesting interview to Fox News on the future of AI and non-US sovereignty in which he insisted:

The hope of the United States is that the pursuit of real AI sovereignty, the adoption and deployment of sovereign infrastructure, sovereign data, sovereign models and sovereign policies within national borders and under national control, will become an occasion for bi-lateral diplomacy, international development, and global economic dynamism.

He went on to dismiss simplistic definitions of what constitutes sovereignty in the first place, stating:

Real AI sovereignty means owning and using best-in-class technology for the benefit of your people, and charting your national destiny in the midst of global transformations. We urge nations to focus on strategic autonomy alongside rapid AI adoption rather than aiming for full self-sufficiency. 

AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralized control. We deeply believe that the best pathway for the developing world to fully realize the untold benefits of AI is through the adoption of the American AI stack.

And there’s the red, white, and blue shining through, no? Actually Kratsios does make a good point that sovereignty isn’t the simplistic soundbite that many pitch it as - that way leads to the ‘sunlit uplands’ of Brexit Britain... - and in this he echoes a point made at Davos by European Aiman Ezzat, CEO of French services giant Capgemini, when he observed:

Sovereignty is not one moralistic thing.

Can countries even afford sovereignty?

It’s a thesis picked up by a study from the Tony Blair Institute for Global Change, entitled Sovereignty in the Age of AI: Strategic Choices, Structural Dependencies and the Long Game Ahead. 

Before pulling out some of the headline conclusions, some caveats - the Institute is bankrolled in part by some mega US tech firms with a lot of skin in the global market, while Blair himself has cultivated the Trump 2.0 administration assiduously, earlier this month popping up on the Board of Peace at a time when the country he used to be Prime Minister of wants nothing to do with it.

So, yes, there will be siren voices who argue that the findings of the Global Institute may be skewed in favor of the US side of the debate. Again, you pays your money, you makes your choice about what to believe.

Nonetheless there are some interesting talking points in the report worthy of further debate, not least the million dollar/euro/pound question of whether countries can actually afford such luxuries as tech sovereignty? The study frames the debate around what it calls two defining questions - what is sovereign AI capability, and how can nations develop and exercise it?

The instinctive  response to that would center around more control over frontier tech and the expansion and consolidation of domestic capabilities, but this, according to the Institute, is an instinct that reflects “a narrow and ultimately counterproductive understanding of sovereignty”.

Such is Blair’s interest in this area - and read into that what you will - that he’s written the foreword to the report himself in which he argues:

Many leaders are now asking what steps they should take to stay in control of their country’s future. The instinctive response is often to try to do everything at home – to build fully “sovereign” AI and treat reliance on partners as a threat.

But that’s wrong, he argues:

Full self-sufficiency is too expensive, too slow and, for most countries, simply impossible. More importantly, it mis-represents what sovereignty really means in a digital, global and interconnected world. Sovereignty should not mean independence from all others. Rather, it should be viewed as the ability to act strategically – with agency and choice – in a world that is irreversibly inter-dependent.

Effective AI sovereignty therefore cannot be pursued through isolation, is the top line conclusion from the report, and governments around the world need to avoid what it pitches as “a futile pursuit of technological self-sufficiency”.

There’s nothing in that conclusion that’s going to see funding from Silicon Valley  for Blair’s thinktank drying up any time soon. In fact the report goes further, explicitly identifying - with justification - that the US and China are the two poles of AI power, but siding clearly with the US axis when it says:

The US promotes a “trusted” AI ecosystem that is rooted in innovation, scale and private enterprise, but reinforced through an industrial strategy that combines subsidies with export controls to secure national security and strategic advantage. China advances a state-directed model that treats AI as a strategic public good while embedding ideological alignment into governance and promoting “openness” as a narrative to normalise its tightly controlled ecosystem.

What does the US want everyone else to do?

While the Blair Institute report isn’t telling us anything revelatory in its conclusion that AI is two-horse race between the US and China - although I could find you more than a few Eurocrats who’ll make a grandstanding stab at denying such a reality - in his Fox News interview, Kratsios tries to steer away from that angle:

We do not see this as being about any one competitor.

So what does the US see it as being about? Well, the color chart comes back into play here as he pitches:

This is about the fact that the United States has the best AI technology in the world, and many countries want it in their ecosystems. 

That the raises the question of what the current US administration is doing to convince the rest of the world of this - and it is the rest of the world we’re talking about here, far less than Europe which by now is pretty much a lost cause in terms of altering existing opinions on this subject.

The answer is, quite a lot, notably:

• The American AI Exports Program’s National Champions Initiative, whereby the US  Commerce Department says it will incorporate partner nations’ leading AI companies into customized American AI Export stacks, this apparently demonstrating that American technology directly strengthens and builds domestic AI capabilities.

• Setting up the US Tech Corps, a new initiative of the Peace Corps to provide volunteer technical talent with import partners to provide last-mile support in deploying powerful AI applications for enhanced public services.

• New international financing in the form of a new fund from the US Treasury Department at The World Bank to help countries overcome AI adoption barriers, joining other financing programs at the Export-Import Bank, the U.S. International Development Finance Corporation, the State Department, and the Small Business Administration.

Some of this will inevitably be viewed with some suspicion by certain parties. Access to US-funded capital, infrastructure and deployment assistance will be tempting for many emerging economies, but, critics will counter, it will risk deepening dependence on a US-controlled tech stack, repeating the same ‘mistake’ that Europe has made.

Détente?

So, what now? Is there any sign of some form of tech détente? Not so much as despite the outreach, the US shows no sign of backing down in its opposition to sovereignty-related pushes by nation states. Last year, US Secretary of State Marco Rubio told EU diplomats to stir up opposition to the Digital Services Act, much hated by US tech vendors. 

And the push goes on - as this article was being published, Reuters broke a story that Rubio has now sent an 'action request' telling his people to fight local data sovereignty initiatives in their locales. Such regulation, he says, "disrupt global data flows, increase costs and cybersecurity risks, limit Artificial Intelligence (AI) and cloud services, and expand government control in ways that can undermine civil liberties and enable censorship."

But at the recent AI Summit in India, Kratsios said that non-US countries need to wake up to the reality of the maxim that AI is changing everything on a global basis and to accept a new approach to the sovereignty debate:

It does not mean waiting to participate in an AI-enabled global market until you have tried and failed to build full self-sufficiency. Complete technological self-containment is unrealistic for any country, because the AI stack is incredibly complex. But strategic autonomy alongside rapid AI adoption is achievable, and it is a necessity for independent nations.

And he promised: 

America wants to help.

My take

The gold standard in AI is made in America.

That’s the bottom line from the Washington side of the debate and there will be no deviation from that worldview.

It’s one that will not convince US detractors or deter their pursuit of self-sufficiency. This constituency might take comfort from one other conclusion of the Tony Blair Institute Report:

No state can dominate every layer of the AI stack.

But the reality is, one is going to give it a damn good try!