AI Is Becoming a Matter of Statecraft

Artificial intelligence was once discussed mainly as a technological revolution. Today, it is increasingly becoming a matter of trade, national security, strategic resources and global power. That shift was evident on Sunday, when US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng began talks in New York covering trade, tariffs, critical minerals and artificial intelligence. The discussions include possible safeguards for advanced AI systems and issues surrounding open- and closed-weight models. They are also taking place against the backdrop of negotiations over Chinese rare-earth magnets and other critical minerals. 

The significance extends far beyond another round of US-China negotiations. It signals that AI has moved from the laboratories of technology companies into the architecture of international relations.

For years, the dominant questions were: How powerful will AI become? How many jobs will it change? How can businesses use it?

The questions are now becoming more strategic:

Who will control AI? Who will regulate it? Who will have access to advanced computing? And who will define the rules governing increasingly capable systems?

These questions are particularly important because AI does not exist only in software. Advanced AI depends on semiconductors, data centres, electricity, telecommunications infrastructure and critical minerals. The technology may appear to operate in the cloud, but that cloud rests on a very physical global supply chain.

The emerging US-China discussions therefore reveal a larger reality: the AI race is simultaneously becoming a race over chips, computing power, energy, minerals and technological influence.

The challenge is not only geopolitical.

Leading AI companies and researchers are themselves debating how increasingly powerful systems should be governed. Recent reporting has highlighted disagreements among technology leaders over regulation, oversight and the appropriate balance between innovation and safety. 

This raises a fundamental question:

Can companies developing the most powerful AI systems also be expected to determine the boundaries within which those systems should operate?

Governments will inevitably have a role. But national regulation alone will be difficult when AI models, data, computing infrastructure and their users operate across borders.

An AI system developed in one country can be accessed in another. An open-weight model can be modified and redistributed. Cybersecurity risks can cross jurisdictions instantly.

International cooperation is therefore necessary.

Yet this is also the paradox of the AI age: the countries whose cooperation may be most necessary are also competing most intensely for technological leadership.

What does this mean for Pakistan?

For Pakistan, this should not be viewed as a distant contest between major powers.

AI could influence almost every area of national development — education, agriculture, healthcare, water management, manufacturing, financial services, energy and public administration.

But there is a danger that Pakistan becomes primarily a consumer of AI technologies developed elsewhere.

Buying access to foreign AI platforms is not the same as developing national AI capability.

Pakistan needs researchers, engineers, entrepreneurs, policymakers and educators who understand both the possibilities and limitations of the technology.

Universities have a particularly important role.

AI education should not be restricted to computer science. Future doctors, engineers, economists, teachers, lawyers and public administrators will all work in environments shaped by AI.

At the same time, universities should become places where AI is applied to Pakistan’s own problems: improving agricultural productivity, managing scarce water resources, strengthening disaster preparedness, developing advanced materials, improving energy systems and supporting public services.

The objective should be to move from AI consumption to AI capability.

That requires investment in research, computing infrastructure, digital skills, responsible governance and collaboration between universities, industry and government.

There is another dimension that deserves attention.

The global conversation about AI is increasingly dominated by a race for capability: bigger models, more computing power and greater automation.

But capability alone cannot be the measure of progress.

A country may have powerful AI systems and still struggle with questions of trust, accountability, privacy, employment and inequality.

The real challenge is therefore institutional.

Can societies build governance systems capable of keeping pace with technologies that are advancing extraordinarily quickly?

For Pakistan, the answer should begin with preparation rather than imitation.

We do not need to reproduce Silicon Valley or compete directly with the world’s largest technology powers. We need to identify the areas in which AI can produce the greatest national value and build the human and institutional capacity to use it responsibly.

That means treating AI not simply as an IT-sector opportunity but as a national development issue.

The current US-China discussions offer an important lesson.

Artificial intelligence is no longer merely about machines becoming smarter.

It is increasingly about who controls the infrastructure, resources, standards and institutions surrounding those machines.

The countries that understand this early will be better positioned to shape the emerging economy.

For Pakistan, the choice is becoming clearer: remain a passive consumer of technologies developed elsewhere, or invest in the people, research institutions and governance frameworks that can make the country an active participant in the AI economy.

The AI era is already here.

The more important question is not whether Pakistan will be affected by it.

It is whether Pakistan will help shape what comes next.