Middle powers may lack the computing infrastructure and frontier laboratories of the United States and China, but they still have significant tools to influence the future of artificial intelligence. Recent debates surrounding the India AI Impact Summit have highlighted a growing strategy centred on regulatory coordination, market access, technical standards, coalition diplomacy and sovereign AI capabilities. Analysts argue that countries such as India, Japan, South Korea, the Netherlands, Germany and other technologically capable states could collectively shape the rules governing AI, even if they cannot compete directly with the world’s two leading AI powers.
Why Are Middle Powers Becoming More Important In AI Governance?
The global AI landscape is increasingly dominated by the technological and financial capabilities of the US and China. Both countries host major frontier AI companies, possess extensive computing infrastructure and have substantial access to advanced semiconductor technologies.
For middle powers, however, influence does not necessarily depend on building the world’s largest AI model. Their leverage can come from regulation, consumer markets, specialist industries, research institutions, skilled workforces and international alliances.
The February 2026 India AI Impact Summit provided an important example. Hosted in the Global South, the summit shifted part of the international discussion towards AI deployment, development, public services and inclusive growth rather than focusing exclusively on frontier-model safety. The International Institute for Strategic Studies described the event as a significant change in the direction of global AI governance.
That shift matters because AI is being adopted in countries with very different levels of infrastructure, income, language diversity and institutional capacity. Rules designed solely by the countries developing frontier systems may not adequately address those differences.
What Can Middle Powers Actually Do?
The most immediate opportunity is collective regulatory and market leverage.
Individual middle powers may have limited influence over a major AI developer. A coordinated group, however, can establish common requirements for systems entering their markets. These could include safety testing, transparency requirements, procurement standards, certification procedures and rules governing high-risk applications.
Such coordination could create commercial incentives for AI companies to meet common standards across multiple jurisdictions.
This approach does not require middle powers to replicate the enormous computing investments being made by the US and China. Instead, it uses something they already possess: access to substantial consumer and business markets.
Researchers have also pointed to coalition diplomacy as another important instrument. Countries with complementary strengths can cooperate rather than attempting to develop every component of the AI ecosystem independently.
How Could AI Coalitions Give Smaller Countries More Influence?
Coalitions can combine capabilities that are distributed across different countries.
The semiconductor supply chain provides a clear example. The Netherlands has a critical role in advanced lithography equipment, Japan is a major technology and manufacturing power, South Korea is central to memory-chip production, Taiwan plays a pivotal role in semiconductor manufacturing, while India has a large pool of technical talent.
No single middle power controls the entire ecosystem. Collectively, however, these countries possess assets that are difficult for the major powers to ignore.
This creates an opportunity for governments to coordinate positions on technology standards, supply-chain resilience, AI safety and market access.
The strategy resembles traditional middle-power diplomacy: rather than attempting to dominate an issue, countries use specialised capabilities and partnerships to increase their bargaining power.
Can Middle Powers Build Their Own Sovereign AI?
Another option is investment in sovereign AI capabilities.
Sovereign AI generally refers to a country’s ability to develop, control or deploy AI systems in ways that reflect its own national interests, rather than relying entirely on foreign providers. Chatham House has identified sovereign AI strategies as one possible response to the dominance of the US and China.
For middle powers, this does not necessarily mean constructing a frontier model capable of matching the most advanced systems.
Instead, governments can concentrate on strategically important areas such as domestic computing infrastructure, national datasets, language models, AI safety testing, research capacity and public-sector applications.
India, for example, has sought to develop AI capabilities suited to its large and linguistically diverse population. Its approach has placed considerable emphasis on practical deployment, including applications in health, agriculture, education and public services.
Such investments can also reduce strategic vulnerabilities. Dependence on foreign AI infrastructure could become a problem if geopolitical tensions disrupt access to computing resources, software, models or technical expertise.
Why Is Regulation Becoming A Strategic Tool?
AI regulation is no longer simply a question of consumer protection or technology policy. It is increasingly becoming part of economic and geopolitical strategy.
Countries that establish credible standards for AI testing, transparency and safety can influence how international companies develop and deploy their products.
The emerging model is not necessarily based on one global treaty. Instead, common practices can gradually develop through cooperation between regulators, technical organisations and governments.
The International AI Safety Report, for example, has helped establish an evidence-based approach to assessing rapidly changing AI capabilities. The report involved more than 70 international experts and received backing from more than 30 countries, according to Techplomacy Magazine.
Experts argue that voluntary commitments can eventually develop into shared assessment methodologies and common standards. Once enough markets adopt similar expectations, companies have stronger incentives to follow them globally.
Could US-China Rivalry Make Middle Powers More Powerful?
The rivalry between Washington and Beijing presents both a threat and an opportunity.
The principal danger is fragmentation. If AI governance develops into two incompatible systems, countries could be pressured to choose between American and Chinese technological ecosystems.
That could make it harder to establish common safety standards, incident-reporting mechanisms and technical evaluation procedures.
At the same time, rivalry between the two largest AI powers creates space for countries that can maintain dialogue across competing blocs.
Neutral or broadly trusted forums could provide channels for technical discussions when direct US-China diplomacy becomes difficult. Researchers argue that these mechanisms cannot replace agreement between major powers, but they can keep international coordination functioning.
This makes diplomatic flexibility an important asset for middle powers.
What Are The Biggest Obstacles To This Strategy?
The greatest challenge is coordination.
Middle powers have different economic interests, security relationships and domestic regulatory systems. A country closely aligned with Washington may approach Chinese technology differently from a state seeking strategic autonomy.
There are also significant differences in technological capacity. Not every middle power possesses advanced semiconductor industries, large research institutions or substantial domestic computing resources.
Another problem is speed. AI development is moving faster than many governments can design and implement new rules.
The rapid development of agentic AI is particularly challenging. These systems can perform sequences of tasks with limited human intervention, creating difficult questions about accountability when autonomous actions cause harm. Existing governance frameworks were largely designed around humans initiating actions and software executing instructions.
Fragmented regulation can also increase compliance costs without necessarily improving safety, particularly for smaller companies and developing economies.
How Could AI Safety Become A Shared Resource?
One proposed solution is greater international cooperation on AI safety infrastructure.
Instead of every country developing separate testing systems, governments and research institutions could share evaluation methodologies, red-team techniques, risk-assessment frameworks and monitoring tools.
This could reduce duplication while allowing countries with fewer resources to participate in international AI safety efforts.
The concept has been compared with public-health infrastructure, where common testing and certification standards can be adapted to local conditions.
For middle powers, such cooperation could provide both practical and diplomatic benefits. Countries would gain access to technical expertise while helping shape the standards that govern AI deployment.
What Happens Next For Middle Powers?
The next phase of AI governance is likely to be less about creating one grand international institution and more about developing practical cooperation between governments, regulators, researchers and technology companies.
The sequence of AI summits from Bletchley in 2023 to Seoul in 2024, Paris in 2025 and India in 2026 illustrates how the debate has expanded from frontier risks towards implementation, development and broader participation.
For middle powers, the opportunity is therefore clear but limited. They are unlikely to displace the US or China as the dominant forces in frontier AI. Their more realistic objective is to prevent those two powers from defining the entire governance system without wider participation.
That means investing selectively in domestic capabilities, coordinating regulations, protecting critical technology supply chains and building coalitions around shared standards.
The outcome will depend on whether these countries can turn individual strengths into collective leverage before the global AI architecture becomes permanently divided. As AI systems become more capable and more deeply embedded in economies and public services, the countries that help establish practical standards today could have considerably more influence over the technology’s future than their individual size might suggest.
