The convergence of artificial intelligence, commercial space infrastructure, and intensifying great-power competition has, in the first four months of 2026, produced what may be remembered as the inflection point of the orbital century. Three developments stand out. First, capital is moving into space infrastructure at a pace the sector has never seen. Second, the architecture of artificial intelligence itself is beginning to migrate off-planet. Third, the doctrines that govern how states behave in orbit are being rewritten in real time, with the United States, China, and Russia each accelerating counterspace programs that were, only a few years ago, treated as theoretical.
For policymakers, investors, and operators across finance, defense, and technology, the signal is consistent: orbital assets are no longer a strategic complement to national power. They are increasingly its operational foundation.
The Capital Surge
Global investment in space infrastructure more than doubled year over year in the first quarter of 2026. According to Space Capital, $6.7 billion was deployed into hardware and software for building, launching, and operating orbital assets in the quarter alone, the third largest quarter on record and a pace that puts 2026 on track to exceed the $55.3 billion raised across the sector in 2025. Across all segments tracked, $36 billion was invested in the quarter.
The capital is following a clear thesis. Space is no longer treated by institutional allocators as a long-duration moonshot category. It is increasingly framed as a near-term industrial buildout, with four themes pulling investor attention: AI and space, national security convergence, lunar economics, and the commercial low-Earth-orbit transition. Sitting above all of them is the prospect of a SpaceX initial public offering at a valuation north of one trillion dollars, an event that, when it arrives, is expected to force every major institutional fund to take a position on the orbital economy whether or not it has previously done so.
The broader market context reinforces the thesis. The global space economy, valued at roughly $600 billion in 2024, is projected to approach $1.8 trillion by 2035. That growth is no longer driven primarily by government programs. Commercial revenues, particularly in downstream services such as Earth observation, navigation, and data analytics, now account for the majority of sector expansion. Governments, in turn, are increasingly dependent on private operators for capabilities once considered sovereign functions.
Artificial Intelligence Moves Off-Planet
The most consequential structural shift of 2026 is the migration of compute itself into orbit. For most of the AI era, the question of where models are trained and run has been an Earth-bound debate about power grids, water consumption, land permitting, and community opposition to hyperscale data centers. That debate is now being challenged by a credible, well-capitalized alternative.
Starcloud, a Redmond, Washington startup, became the fastest Y Combinator company in history to reach unicorn status, closing a $170 million Series A at a $1.1 billion valuation in late March. The round was led by Benchmark and EQT Ventures, with participation from infrastructure investors that operate large terrestrial data center portfolios. The company has now raised $200 million in total. In November 2025, Starcloud launched its first satellite carrying an Nvidia H100 chip and reported the first training of a large language model in orbit. A second satellite, with roughly one hundred times the power generation of the first and built around Nvidia’s Blackwell B200 chip, is scheduled for launch later this year.
Starcloud is not alone. Google has launched Project Suncatcher, a long-range plan to deploy tensor processing units in orbit in partnership with Planet Labs. Aetherflux, Axiom Space, Kepler Communications, and Sophia Space are pursuing variations of the same model. Nvidia has formalized its position as the underlying compute provider for the segment, announcing in March its Space-1 Vera Rubin Module, which the company says delivers up to twenty-five times more AI compute for space-based inference than its H100 baseline.
The most aggressive move has come from SpaceX itself. In January, the company filed an application with the Federal Communications Commission for the deployment of up to one million satellites configured for orbital data center operations, a filing that, if approved, would represent an order-of-magnitude expansion over the roughly fourteen thousand active satellites currently in orbit. Elon Musk has framed the rationale in terms of physics rather than politics: terrestrial AI is power-constrained, and orbit offers continuous solar irradiance and an effectively infinite cold sink for waste heat.
The economics, however, remain unsettled. Current launch costs run roughly $1,000 to $3,000 per kilogram. Industry analysis, including a feasibility study published by Google in late 2025, suggests orbital data centers become cost-competitive with terrestrial facilities when launch costs fall below roughly $200 per kilogram. That threshold depends on SpaceX’s Starship achieving operational reusability and a launch cadence that most independent analysts do not expect before the early 2030s. Until then, the segment will operate as a high-conviction infrastructure bet predicated on a single company’s engineering execution.
For Intelligence Report’s audience, the strategic implication is clear. If orbital compute scales as projected, the geography of artificial intelligence will shift from the American Southwest, the Nordic data corridor, and the Persian Gulf into low Earth orbit, where jurisdiction, taxation, energy regulation, and physical access become genuinely contested questions for the first time.
Counterspace: The Doctrine Catches Up With the Hardware
While capital flows into commercial orbital infrastructure, the security architecture surrounding those assets is being rebuilt in parallel. The Secure World Foundation’s 2026 Global Counterspace Capabilities Report documents active counterspace programs in thirteen countries, with development now spanning five categories: co-orbital systems, direct-ascent anti-satellite weapons, electronic warfare, directed energy, and cyber operations.
The United States has signaled a doctrinal shift. The 2026 National Defense Strategy elevates space as a foundational enabler of homeland defense and Western Hemisphere security, with surveillance constellations explicitly tied to monitoring the Panama Canal, Greenland, and maritime approaches across the Caribbean. The Space Force’s Race to Resilience initiative targets battle-ready architectures by year-end 2026, supported by record fiscal year funding approaching $40 billion. Four on-orbit servicing demonstrations, covering refueling, repair, inspection, and maneuvering, are scheduled for the year. The Commercial Augmentation Space Reserve, which gives the Pentagon contractual access to private satellite networks during conflict, will transition from pilot to full-scale operations with a target of twenty contracts by year-end.
Above all of this sits the Golden Dome program, the Pentagon’s missile defense architecture, which is on schedule for an operational capability by 2028. A consortium of nine defense firms recently completed a live demonstration of the system’s command-and-control layer. Under Defense Secretary Pete Hegseth, the program has been paired with public statements about deploying space-based interceptors capable of neutralizing ballistic missiles, hypersonic weapons, and drones before they reach U.S. territory.
China has continued the buildout that the U.S.-China Economic and Security Review Commission described in its November report as positioning Beijing to displace the United States as the world’s premier space power. China’s operational satellite fleet exceeded 1,060 spacecraft by mid-2025, with several hundred dedicated to intelligence, surveillance, and reconnaissance. Beijing’s approach to counterspace, as documented in academic and government analyses, has prioritized soft-kill capabilities, jamming, spoofing, lasers, and cyber operations, that can be employed early in a crisis with lower escalation risk than kinetic anti-satellite strikes.
A Space Force assessment titled Future Operating Environment 2040, published earlier this month, projects that by the end of the next decade China will be conducting sustained low-level space competition powered by AI-enabled decision systems, brain-computer interfaces, and metamaterials including what the report describes as satellite cloaking technologies. The same report describes a Russian doctrine built around maneuvering robotic satellites, mobile jammers, and dormant orbital systems designed to activate at unexpected moments, accompanied by what U.S. analysts call “repeated accidents, falsified ephemerides, and escalatory nuclear signals” intended to complicate attribution and crisis management.
In late April, congressional testimony from the Center for Strategic and International Studies disclosed that a Chinese AI firm, MizarVision, has publicly demonstrated the ability to track U.S. military assets and movements using satellite imagery, while Iran has acquired a Chinese imaging satellite and Russia has supported Iranian operations with its own overhead data. The disclosure illustrates a pattern that will define the next decade: AI-processed commercial imagery, sold or shared across hostile coalitions, increasingly performs functions that were once the exclusive domain of national intelligence services.
Governance: Form Without Substance
The governance architecture surrounding these developments has not kept pace. The 1967 Outer Space Treaty, the foundational legal framework for orbital activity, contains substantial ambiguities around in-orbit servicing, dual-use payloads, mega-constellations, and the question of what constitutes a use of force in space. Wealthier states have proven able to operate freely within those ambiguities.
The Atlantic Council’s January 2026 forecast on AI and geopolitics noted that international cooperation on AI governance has converged on shared scientific assessments and voluntary principles but has consistently avoided binding constraints on the highest-stakes applications, including autonomous weapons, mass surveillance, and information manipulation. The same dynamic is now visible in space governance. A United Nations General Assembly resolution in 2022 supporting a moratorium on destructive direct-ascent anti-satellite testing was endorsed by a number of states, but enforcement remains voluntary and selective.
The result is a governance architecture that is, in the Atlantic Council’s framing, global in form but geopolitical in substance. Norms manage risk at the margins. Strategic competition between rival systems remains structurally unresolved.
Implications for Decision-Makers
Several conclusions follow for leaders operating at the intersection of capital, policy, and technology.
The investment window for early positioning in orbital infrastructure is narrowing rather than widening. The pending SpaceX listing, the Nvidia space hardware ecosystem, and the maturation of orbital data center pilots are likely to compress what would otherwise be a multi-year discovery period for institutional allocators into a far shorter window. Family offices and sovereign funds that have historically treated space as a niche allocation are being forced to revisit the question.
National security convergence with the commercial sector is no longer a forecast but an operational reality. The Pentagon’s Commercial Space Integration Strategy, the Maneuverable GEO program, and the Commercial Augmentation Space Reserve all assume a hybrid military-commercial architecture as the default. Companies that build dual-use capability with deliberate posture toward both customers will face very different regulatory and capital environments than those that do not.
The governance gap will be filled, eventually, but the entities that shape the new norms will be those present in the room at the moment of crisis. The United Nations system, which has historically convened space governance discussions, will remain a critical venue, particularly during the General Assembly cycle and through specialized agencies including the International Telecommunications Union and the Committee on the Peaceful Uses of Outer Space. Convening capacity within and adjacent to these institutions is itself a form of strategic leverage.
Finally, the operational tempo of orbital competition is accelerating beyond the speed at which traditional human-in-the-loop decision frameworks can function. AI-driven space-based decisions occur in microseconds. Doctrines, treaties, and regulatory regimes designed for a world in which decisions could be deliberated in hours or days will require fundamental redesign. That redesign is unlikely to happen in advance of the first significant orbital incident. It will happen in response to one.
The Intelligence Report
The orbital century is no longer a forecast. It is the operational environment in which the next phase of great-power competition, capital formation, and technological development will unfold. The actors and institutions that recognize this early, and that build the relationships, infrastructure, and policy fluency to operate within it, will define the strategic landscape for decades.
For the rest, the window to position is closing.
Intelligence Report is a global media and intelligence company providing analysis at the intersection of geopolitics, capital, and emerging technology. This analysis draws on open-source reporting from Space Capital, Brookings, the Center for Strategic and International Studies, the Atlantic Council, the Secure World Foundation, the U.S. Space Force, SpaceNews, NPR, TechCrunch, GeekWire, the Washington Times, and Eurasia Review.
