AI Sovereignty Is Not Independence

The real strategic question is whether countries and organizations can retain choice when access to technologies they do not control begins to change.

In partnership with

For much of the AI era, access has been treated almost as a constant.

Choose a model. Choose a cloud provider. Integrate it into the organization. Assume that tomorrow's access will look broadly like today's.

That assumption is becoming harder to defend.

Chinese authorities have been discussing whether to restrict overseas access to the country's most advanced AI models, including future models that have not yet been released. Officials have reportedly considered tighter controls around frontier models, stronger penalties for leaking proprietary AI technology and restrictions on who can finance domestic AI companies.

No final policy has been announced. The scope remains unsettled.

But that uncertainty is precisely the point.

A government decision in Beijing could change the technology choices available to companies and countries thousands of kilometres away - without any of them having a role in making it.

Europe is approaching the same challenge from another direction. Nineteen EU member states are developing IPCEI-AI, the first Important Project of Common European Interest focused specifically on artificial intelligence. Alongside it, Europe is pursuing up to seven AI Gigafactories designed to expand the continent's computing capacity.

Africa introduces another dimension.

At GITEX 2026 Nigeria, AfCFTA Secretary-General Wamkele Mene argued that African countries should build their own digital capabilities while ensuring that national systems can still work across borders.

Together, these developments point to a much larger question:

What happens when critical AI capabilities depend on decisions you do not control?

The answer is not complete technological independence. It is strategic optionality.

Some teams never seem to stop moving. They're on Attio, the agentic CRM.

Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.

With Attio, you’ll get:

  • Leads automatically prioritised and routed to the right rep

  • Expansion and risk signals caught the moment they land

  • Follow-ups written in your voice, already there when you arrive

Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?

THE ASSUMPTION UNDERNEATH THE AI BOOM

Most organizations do not think of their AI architecture as geopolitical infrastructure. They probably should.

Reuters reported that Chinese officials had been discussing with companies including Alibaba, ByteDance and Z.ai whether to restrict overseas access to China's most advanced AI models.

The discussions reportedly extend across both closed and more open models. Officials have also considered stronger national-security protections for proprietary AI technology and additional restrictions around foreign financing of domestic AI startups.

Whether those proposals become policy remains uncertain. Their significance does not.

Chinese models have become increasingly attractive internationally as their capabilities improve and costs remain competitive. At the same time, the United States already uses export controls to restrict access to strategically important technologies, particularly advanced semiconductors.

AI access is becoming increasingly conditional. And that changes the meaning of diversification.

An organization using technologies originating in several countries may appear diversified from a vendor perspective while remaining highly concentrated from a geopolitical one.

The model may come from one country. The cloud from another. The advanced chips from a third company. The fabrication capacity may sit somewhere else again.

What looks like a technology stack is increasingly also a geopolitical supply chain.

That is the strategic issue.

EUROPE IS BUILDING STRATEGIC CAPACITY

Europe's response illustrates an important distinction between sovereignty and resilience.

Nineteen EU member states are developing IPCEI-AI, coordinated by Germany, to create advanced and more widely accessible European AI capabilities across the AI value chain.

The initiative sits within a much larger effort.

The European Union has launched a tender for up to seven AI Gigafactories, supported by as much as €10 billion in EU and national funding and intended to mobilize at least another €20 billion in private investment.

Meanwhile, European AI companies are attracting substantially more private capital. Mistral AI raised €3 billion in September at a valuation above €21 billion.

Europe is expanding the portion of the AI value chain over which it has greater strategic control.

But Europe's experience also reveals the limits of the word sovereignty.

The European Commission has signed letters of intent with AMD, Nvidia and Qualcomm to help participating Gigafactory consortia access advanced hardware.

Europe may build European models, European infrastructure and European governance while still depending on semiconductor technologies and global manufacturing supply chains beyond its direct control.

That does not make the strategy less meaningful. It makes the objective clearer.

Strategic autonomy does not require producing everything yourself. It requires ensuring that losing access to one supplier, technology or jurisdiction does not disable your ability to act.

Europe is not eliminating dependency. It is attempting to make dependency manageable.

Learn AI in 5 minutes a day

You don't have to scroll every AI thread, track every new tool, or watch every demo. 

The Rundown AI breaks it all down for you — the latest AI news, tools, and tutorials in one free 5-minute email every morning. 

Trusted by 2M+ professionals at Apple, Google, and NASA.

AFRICA ADDS THE MISSING QUESTION

There is another risk in the pursuit of technological sovereignty.

Countries can become so focused on reducing external dependency that they create fragmentation among themselves.

At GITEX 2026 Nigeria, AfCFTA Secretary-General Wamkele Mene argued that Africa should not simply import digital infrastructure designed elsewhere. The continent has the talent and technical capability to shape its own digital transformation.

But the more consequential question is what comes next.

African digital capability must still operate across African borders.

The AfCFTA Protocol on Digital Trade seeks to create common approaches around digital infrastructure, electronic authentication, digital identities, payments and cross-border data flows.

That matters because 54 sovereign digital systems that cannot communicate with one another would reproduce digitally many of the barriers African economic integration is attempting to remove physically.

This is where the African and European stories intersect.

Sovereignty without interoperability can become isolation.

A country can control its technology and still discover that the technology cannot participate effectively in the markets around it.

The same is true for companies.

A proprietary AI environment that cannot communicate with customers, suppliers, regulators or other platforms may technically provide control while economically reducing flexibility.

The objective should therefore not be independence at any cost. It should be control where it matters and interoperability where it creates value.

THREE QUESTIONS EVERY EXECUTIVE SHOULD ASK

The geopolitical dimensions of AI can sound distant from ordinary enterprise decisions. They are not.

Every executive team should understand three things about its AI environment.

Where are we dependent?

Map the models, cloud infrastructure, chips, APIs and critical platforms underneath the organization's AI capabilities. Knowing the name of the application is no longer enough. Leaders need to understand what sits beneath it and which jurisdictions ultimately influence access.

Where do we have optionality?

The objective is not to eliminate every external dependency. That is unrealistic for most organizations - and almost every country. The question is whether losing access to one provider or technology would create inconvenience or paralysis. Those are very different risk positions.

Can what we build still connect?

Resilience should not come at the expense of interoperability. National systems must participate in international markets. Enterprise systems must work across customers, partners and regulatory environments. Optionality without interoperability simply replaces one dependency with another constraint.

AI SOVEREIGNTY IS REALLY ABOUT CHOICE

China has not announced the restrictions currently under discussion.

Europe has not achieved technological independence.

Africa has not solved the challenge of building an interoperable continental digital market.

That is precisely why these developments matter now.

They reveal the direction of travel before the destination is fixed.

For years, much of the AI conversation has focused on capability: who has the most powerful model, the largest compute cluster or the most advanced application.

The next phase will increasingly be about control, dependency and choice.

Countries and organizations do not need to control every layer of the AI stack.

But they should know which layers they do not control. They should understand what happens if access changes. And they should be building alternatives before circumstances force them to.

AI sovereignty is not independence.

It is the ability to retain strategic choices when access to technology you do not control changes.

In the emerging AI economy, the defining advantage may not be owning everything. It may be retaining the freedom to choose.

Discover Bridging Worlds, a thought-provoking book on technology, leadership, and public service. Explore Lawrence’s insights on how technology is reshaping the landscape and the core principles of effective leadership in the digital age.

Order your copy today and explore the future of leadership and technology.

SHARE YOUR THOUGHTS
We value your feedback!

Your thoughts and opinions help us improve our newsletter. Please take a moment to let us know what you think.

How would you rate this newsletter?

Login or Subscribe to participate in polls.

Reply

or to participate.