The global artificial intelligence investment cycle is likely to extend as governments prioritize sovereign control over AI infrastructure, data and technology, according to a Morgan Stanley research note published on August 23, 2026.
The U.S. is expected to bolster domestic AI capabilities while tightening restrictions on advanced technology flows to China, creating a bifurcated market for AI compute, models and cloud services. Morgan Stanley’s base case assumes a balanced approach that safeguards frontier AI intellectual property while maintaining openness where feasible.
China is projected to pursue localization through domestic procurement, cybersecurity regulations, local model integration and sovereign cloud infrastructure. The strategy aims to mitigate Western market restrictions by expanding into emerging markets using lower-cost, open-source technologies and strategic partnerships.
In Europe, efforts to achieve strategic autonomy in compute, cloud and semiconductors face financing and scale constraints. Meanwhile, ASEAN countries are rapidly developing domestic AI infrastructure and forming alliances with trusted international partners, positioning the region as one of the fastest-growing opportunities in sovereign AI deployment.
Morgan Stanley estimates U.S. AI capital expenditure will reach $860 billion in 2026, underscoring the scale of investment driven by both public and private sectors. The push for sovereign AI is expected to reinforce spending across semiconductors, data centers, networking equipment, power infrastructure, cloud services and AI deployment platforms.
While AI adoption is projected to enhance productivity in exposed industries, potential risks include task-replacement unemployment, cyclical infrastructure overinvestment and rising inequality. The report highlights that government-led strategic investments may help sustain the AI capital expenditure cycle despite near-term moderation in private sector spending.
