
Artificial intelligence investment is proving less borderless than the language of global technology competition often suggests. New research into more than 1,200 AI supply-chain companies shows that dealmaking is concentrated heavily in home markets, particularly in the largest ecosystems where capital, talent, customers and credible targets are already available.
Between 2010 and 2025, two-thirds of investment deals by nearly 700 US-based AI-producing companies were domestic. Chinese AI firms showed an even stronger domestic concentration, with 74% of deals taking place inside China over the same period. The same home-market bias was also identified in Japan, the UK, the euro area and South Korea, while Canada, India and Sweden stood out as exceptions, with firms there directing more investment towards larger external markets.
The pattern reflects the uneven geography of AI capability. The US accounted for almost 700 of the 1,254 public and private AI companies analysed, across compute, cloud infrastructure, data tools, models and applications. China had 244 firms, while the next-largest ecosystems were far smaller, including Israel, the UK, Japan, Canada, India, Taiwan and Germany. In thinner ecosystems, outward investment becomes more important because firms may struggle to find all the capabilities they require locally.
AI mergers and acquisitions also show a strong domestic tilt compared with other sectors. Domestic acquisitions of AI companies accounted for just under 80% of total AI-targeted deals between 2010 and May 2026, exceeded only by real estate and retail. Sovereignty concerns and the race for AI dominance are helping to keep capital close to national champions.
The unresolved issue is whether countries outside the main hubs can attract meaningful AI capital without sharper positioning. Low costs and broad incentives may not be enough where existing centres are already compounding their advantages through talent density, procurement and sovereign-AI policy.