Middle Powers Build AI Alliance Outside US-China Duopoly

Bilateral Deals Replace Frontier Ambitions

Countries across the Gulf and Southeast Asia are sidestepping the race to develop cutting-edge AI chips and models, opting instead for a network of bilateral infrastructure and investment agreements. The approach leverages what each side already possesses: Gulf capital and energy resources meet Southeast Asian demand and digital scale.

Abu Dhabi's G42 signed a framework deal in February worth up to $1 billion with a Vietnamese consortium to deploy national AI and cloud systems across government, industry, and universities. Malaysia drew $23 billion in data-center investment last year and maintains formal cooperation agreements with the UAE on AI and digital economy initiatives. Masdar, Abu Dhabi's state-owned renewable energy entity, committed $15 billion in January 2025 to build clean power capacity in the Philippines.

These transactions follow a pattern: specific capability matched to specific need, executed without requiring consensus across entire regions. That structure has allowed partnerships to move quickly while global AI governance splinters into incompatible regulatory zones.

Concentration at the Frontier Leaves Room Below

Frontier AI development remains tightly held. Eighty notable AI models launched last year came from just two countries. The United States operates 5,427 data centers, more than ten times the count of any other nation. Among the 33 countries hosting public-cloud AI infrastructure, only two run it on domestically produced chips with domestic providers.

That concentration leaves middle powers dependent on foreign technology imports, but it also opens space for partnerships built on deployment rather than invention. Southeast Asia's digital economy is on track to surpass $300 billion in gross merchandise value by 2025, according to industry projections. The region's data-center market is forecast to grow from $13.7 billion to $30.5 billion by 2030.

Vietnam has enacted standalone AI legislation and aims to position itself as a regional hub; it already hosts over a quarter of Southeast Asia's generative-AI start-ups. Thailand participates in Project mBridge, a cross-border digital-currency platform that also includes the UAE and Saudi Arabia.

Capital Flows Meet Infrastructure Gaps

Saudi Arabia is directing $100 billion toward AI start-ups and data centers under Project Transcendence. The UAE has allocated $148 billion to AI infrastructure over the past two years. Stargate UAE, a 1-gigawatt computing cluster being built by G42 with OpenAI, Oracle, NVIDIA, SoftBank, and Cisco, carries an estimated price tag near $30 billion.

G42's Jais 2 model, trained on 600 billion Arabic tokens, is part of an effort to build Arabic-language AI capacity domestically. The speed and scale of Gulf infrastructure projects create leverage in partnerships with countries that need computing capacity but lack the capital or energy resources to build it themselves.

Singapore, Malaysia, and Vietnam have emerged as primary Gulf partners in Southeast Asia. The Joint Declaration on Economic Cooperation between ASEAN and the GCC explicitly endorsed AI and digital economy cooperation, giving political cover to deals that were already moving.

Pilots Over Platforms

Research from the AI Asia Pacific Institute identifies recurring structures in UAE partnerships with Singapore, Malaysia, and Vietnam: memoranda on digital government, cooperation frameworks under initiatives like Malaysia's MADANI AI, and sectoral agreements that test AI in narrow domains rather than attempting full-stack development.

Southeast Asia's large underbanked populations and expanding fintech sector make it a natural testing ground for AI applications in financial inclusion and small-business credit. The Gulf's legal digitization and administrative capacity position it to pilot AI in courts and public services.

Sectoral pilots allow both sides to generate practical lessons without waiting for comprehensive regulatory alignment. The same model could extend to healthcare, logistics, energy management, and language models tailored for Arabic and Southeast Asian languages.

Geopolitical Hedging Through Economic Ties

Recent conflict involving Iran, the United States, and Israel has underscored the risks of concentrated economic exposure. Gulf states have shown resilience, but the episode has reinforced interest in diversifying investment destinations and supply-chain relationships.

Southeast Asia offers a fast-growing digital market with lower geopolitical entanglement than alternatives. For Southeast Asian governments, Gulf investment provides infrastructure funding and energy partnerships that do not come with the regulatory or security conditions attached to deals with Washington or Beijing.

Neither region will achieve self-sufficiency in frontier AI. Both will continue importing chips, models, and cloud services from the United States and China. But the Gulf-Southeast Asia partnership demonstrates that middle powers can shape where and how AI is deployed, even if they cannot control what it is capable of doing. Investment scale, trusted bilateral relationships, and targeted sectoral deployment create influence without requiring technological independence.