Why Asia-Pacific governments are taking different approaches to sovereign AI

ACROSS the Asia-Pacific, sovereign artificial intelligence has moved from concept to commitment faster than almost any technology agenda in recent memory.

Between 2024 and 2025, it has risen from the seventh to the second-highest investment priority for governments in the region, going by research from market intelligence company International Data Corporation (IDC).

The momentum is not confined to this region. Every nation wants sovereign AI capability and compute.

AI has become the new strategic asset. Yet, treating something as strategic does not mean having to own it outright.

Increasingly, governments are discovering that the most critical choice is not whether to pursue autonomy, but where to draw the boundaries.

A more pragmatic model is emerging as a result – operational sovereignty.

Under that idea, the assumption that digital independence demands full-stack authority gives way to a spectrum of deliberate choices: national control where it is essential, co-development with trusted partners where it creates stronger outcomes, and global ecosystems leveraged where they do not compromise national interests.

Four paths to fulfil AI ambitions

While the regional trajectory towards this operational model is broadly consistent, the execution varies considerably across the Asia-Pacific.

Four distinct approaches are emerging, each representing a different but legitimate point on the spectrum of sovereign AI ambition: governance as an accelerator, national self-reliance, sector selectivity and public adoption.

Each government leverages a different process, and where each lands depends not only on ambition, but also on resource realities, institutional maturity and strategic circumstance.

AI governance is the primary layer of sovereignty control for many governments.

By establishing clear frameworks early, nations such as Singapore, Malaysia and the Philippines are able to scale AI adoption while managing risk.

They do so with the goal of building home-grown capabilities that strengthen security and create economic value in strategic sectors.

Singapore exemplifies this approach.

In January 2026, the city-state launched the Model AI Governance Framework for Agentic AI and committed more than S$1 billion to national AI R&D, demonstrating the way robust governance can serve as a catalyst for innovation.

On national self-reliance, countries with resources and scale are emphasising domestic compute. India and Japan, notably, are investing heavily in control over data infrastructure, local large language models (LLMs) and hardware supply chains.

These actions are driven by the need to reduce technology dependence on others, and secure critical national systems from external disruptions.

The IndiaAI Mission, for example, has onboarded more than 38,000 high-end graphics processing units to domestic infrastructure.

Meanwhile, Japan’s sustained investment in local LLMs, through programmes such as those built on the Fugaku supercomputer, ensures that foundational AI capabilities reflect national linguistic and cultural context.

These are not isolationist moves.

They are rational decisions by nations with the industrial base and population scale to justify sovereign compute as a strategic asset, creating the foundation upon which innovation ecosystems, including international partnerships, can operate.

On the sector selectivity front, governments carefully choose which sovereign controls to apply to targeted sectors.

Australia and New Zealand, for instance, concentrate on domains in which system failure or data breaches would carry serious national consequences.

For example, Australia’s 2025 update to its Protective Security Policy Framework introduced new safeguards for AI and other emerging technologies, alongside stricter requirements for critical government systems and digital infrastructure.

As for public adoption, some governments are focusing on the rapid roll-out of AI.

In South Korea, Thailand and Indonesia for instance, the immediate focus is broad public-sector AI adoption to modernise services and improve citizen outreach.

These governments prioritise deployment and apply strict sovereign controls selectively to risk-sensitive workloads.

This is not sovereignty deferred. It is sovereignty informed by operational experience, ensuring that when controls are applied, they target genuine risks rather than theoretical ones.

Sovereignty through alignment

In the region, what is taking shape among governments is an understanding that being the leader in AI increasingly means knowing what is worth controlling and what is better built together.

The computing layers, data centre architecture and platforms underpinning modern AI represent decades of engineering refinement and capital investment.

Most governments can hardly replicate this at the speed the moment demands and, increasingly, the consensus in the region is that they do not need to.

Hence, many Asia-Pacific governments are now pursuing operational and selective sovereignty models, maintaining strong control over sensitive data, critical systems and regulated workloads, while continuing to leverage global technology ecosystems for innovation and scale.

IDC’s research reinforces this – governments are evaluating vendors less on stand-alone products and more on long-term national partnership capability, with ecosystem participation, responsible AI maturity and knowledge transfer now outweighing traditional product differentiation.

By anchoring infrastructure through trusted commercial relationships, they free up the capacity to concentrate their efforts on what cannot be outsourced.

This includes defining how AI is governed, shaping how it integrates into economic and civic life, and ensuring that the benefits of adoption are captured nationally.

What separates the most resilient approaches, however, is not the structure of individual partnerships, but how adaptable the broader ecosystem is.

Agentic systems and shifting hardware bottlenecks are already redrawing the landscape faster than policy cycles can follow.

The ability to detect such technological shifts and respond decisively depends on how well the ecosystem is organised – how the public and private sectors share signals, coordinate standards and move in step as the technology evolves.

This alignment is what transforms infrastructure investments into a living, responsive strategic asset, and what will make sovereign AI a functioning national infrastructure.

The writer is president for Asia-Pacific, Japan and Greater China at Dell Technologies