2025 became a turning point for global foreign direct investment. Despite declining overall FDI volumes worldwide and muted dynamics in traditional cross-border M&A, the year consolidated a new investment model: fewer deals, larger scale; more capital, concentrated in fewer strategic directions.
Global capital is no longer “evenly distributed.” Investors are increasingly less willing to diversify geographically for diversification’s sake. Instead, investment is becoming tightly concentrated in the core nodes of the future economy—computing, energy, semiconductors, industrial automation, defense and dual-use technologies.
In effect, 2025 marked a new wave of industrialization—yet this time its core is not mass-production factories, but infrastructure for the digital economy.
AI as a Physical Industry: Data Centers, Energy, and a New Class of Capital
The most visible transformation of 2025 was the shift of artificial intelligence from the software environment into the physical economy. AI stopped being “just code and models.” It became a consumer of electricity, land, water, engineered networks, permits, and heavy capex.
That is why, in 2025, data centers definitively moved from the category of “IT real estate” into a full-fledged industrial asset class—comparable in complexity and capital intensity to heavy industry.
The world’s largest technology companies began integrating computing and energy into a single investment logic. Acquisitions of energy and infrastructure platforms, direct investments into generation, long-term power purchase agreements, and hybrid “data center + power” projects became mainstream.
In Europe, one of the most illustrative cases was a large-scale data center campus in Amsterdam with capacity measured in tens of megawatts, which secured an anchor hyperscaler customer already at the investment stage. This highlights the new reality: capital goes where electricity and permits are available before construction begins—not the other way around.
In parallel, a different model is emerging in the U.S.: megaprojects in compute infrastructure with multi-decade horizons and tens of gigawatts of capacity, requiring bespoke financing, special-purpose vehicle (SPV) structures, and syndicated private capital. For the first time in history, data centers are increasingly viewed as foundational economic infrastructure—alongside roads and power grids.
Major 2025 Deals
Alphabet / Google: acquisition of Intersect Power for $4.75 billion
This transaction is a classic example of how tech giants are integrating energy and computing into one chain: Intersect will build power generation assets and AI data centers to support large-scale compute.
$14 billion in European data centers by Goodman and Canada Pension Plan
The joint venture aims to develop multiple data centers in Paris, Frankfurt, and Amsterdam, reinforcing the trend toward geographic diversification of compute capacity on a global scale.
OpenAI & Stargate: approximately $500 billion in planned investments
The Stargate project envisions building a network of data centers with total capacity of roughly 10 GW across the U.S., Latin America, and the Middle East—one of the most ambitious AI infrastructure plans in recent history.
Semiconductors: Investment as Public Policy
The second systemic block of large investments in 2025 was semiconductors. Importantly, the market moved from announcements to engineering execution.
After the wave of declarations in 2022–2023, 2025 shifted the focus to practical constraints: grid interconnection, water availability, materials logistics, workforce readiness, and integration of fabs into local ecosystems.
Investments in fabs, advanced packaging plants, and supporting infrastructure are no longer viewed as purely corporate decisions. They have become part of national economic security strategies. Governments increasingly act as co-investors through subsidies, infrastructure commitments, tax incentives, and guarantees.
At the same time, 2025 showed that not all announced projects are equally viable. Investors began clearly distinguishing political statements from projects that are truly advancing through EPC stages, grid connections, and workforce build-out. This led to revisions in the pace and scale of certain investments—but not a reversal of the trend itself.
Major 2025 Cases
TSMC Arizona: a $165 billion “gigafab” project in the U.S.
One of the largest FDI projects in modern history: a multi-fab complex in Arizona supporting the localization of advanced chip manufacturing.
Qualcomm and strategic acquisitions: Alphawave, Ventana, Arduino
Qualcomm executed several strategic acquisitions in 2025 totaling over $2.4 billion to strengthen its position in AI hardware.
Batteries, EVs, and Energy Value Chains: From Expansion to Optimization
The battery and e-mobility segment entered a maturity phase in 2025. Where earlier years were defined by aggressive capacity expansion, the focus shifted to efficiency, utilization rates, and vertical integration.
Major players revised JV structures, reallocated assets, and optimized supply chains. Investment did not stop—but it became more selective and increasingly tied to real demand, both from automotive markets and grid-scale energy storage.
This is a key signal for 2026: the market no longer funds “capacity for capacity’s sake.” Capital flows to projects backed by long-term offtake contracts, grid integration, and manufacturing flexibility.
Examples from 2025:
-
Toyota and Tesla continued battery manufacturing capex in the U.S. and EU respectively, emphasizing integration of EV and grid-storage solutions.
-
LG Energy Solution sold Honda-related assets for $2.86 billion to optimize production capacity for e-mobility markets.
The State as an Architect of Investment
Another defining feature of 2025 was the sharp increase in the role of governments in the competition for FDI. Investment decisions are increasingly made not solely on IRR.
Investors evaluate the full “jurisdiction package”: tax conditions, permitting speed, regulatory stability, infrastructure access, and a government’s willingness to share risk.
Expanded lists of priority sectors, new incentive catalogs, and special regimes for high-tech and advanced manufacturing have become standard for major economies. In effect, a global race for industrial platforms is emerging—where winners are not those offering cheap labor, but those offering speed and predictability.
What to Watch in 2026: Key Vectors
In 2026, large-scale direct investment is likely to evolve along several tracks.
1) AI infrastructure will remain dominant—shifting toward solar/green power components
The need for stable electricity for AI campuses will drive new hybrid investments combining data centers with generation and energy storage.
2) Geopolitical competition is becoming technological competition
Strategic rivalry over semiconductors, AI, and critical supply chains will shape not only FDI volumes but also export and industrial policy.
3) The new data economy and edge computing
Megaprojects are no longer only about hyperscale hubs—edge localization will become critical for latency-sensitive AI services.
4) “Finance Instruments 2.0”: SPVs and private credit for AI infrastructure
Cases of $120+ billion SPV financing for AI data centers show that traditional banks are no longer the only capital in the game.
5) Localization and subsidies become decisive
Investment catalogs, local incentives, and customs regimes are creating “geographic corridors” of FDI.
6) Resilience and ESG as the primary barriers to entry
Investors will assess not only IRR, but also ESG parameters and the energy footprint of major assets.
7) Regional expansion of AI campuses
Latin America, the Middle East, and Southeast Asia are likely to become the next active platforms—Stargate Argentina’s $25 billion concept supports this trajectory.
Ukraine’s Potential Place in This Architecture
In 2025, Ukraine was not among the primary recipients of large global FDI flows, yet it retained strategic relevance in several areas.
First, this includes engineering and R&D centers in IT, AI, defense, and dual-use technologies. Ukrainian human capital remains a competitive advantage despite elevated risk.
Second, industrial parks and manufacturing sites could become entry points for localization of certain links in European and global supply chains—not as a “factory of the world,” but as a flexible industrial rear base.
Third, energy decentralization, storage solutions, renewables, and resilience infrastructure may become areas where foreign capital is willing to engage—provided risks are structured properly.
The key challenge remains security guarantees, access to long-term financing, and institutional predictability. But in a world where investment increasingly carries geopolitical motivation, even higher-risk jurisdictions can secure a niche—if they articulate it clearly.
Summary
2025 demonstrated that large investments are no longer purely economic decisions. They are strategic bets on the future architecture of the global economy.
AI infrastructure, semiconductors, energy, and industrial technologies are reshaping the capital map. The winners will be those who can integrate engineering, policy, and finance into a single system.
2026 will not be a year of “mass FDI return,” but it may become a year of high-quality investment concentration that defines the economic and technological hierarchy of the next decade.
