On 23 September 2026, a disclosure that had been circulating in hushed industry conversations for weeks became public: SpaceX has been granted access to the Pentagon's most sensitive orbital tracking data, access that no other commercial firm receives on a routine basis 
breakingdefense.com. For founders building in space domain awareness, this is not background noise. It is a direct market event with a permanent effect on the competitive baseline.
The Exchange
The arrangement is bilateral. A contractual agreement has been signed between SpaceX and the Space Force's Joint Commercial Operations Cell (JCO), trading SpaceX's access to US Space Command's most detailed version of its High Accuracy Catalog (HAC) in exchange for data from SpaceX's Stargaze star-tracking system breakingdefense.com. The HAC is the most comprehensive government database of objects in orbit. Its most sensitive layer includes not just positional data but also covariances: volumetric margin-of-error bubbles that allow more precise plotting of potential collisions. Those error bubbles have historically been classified and largely kept for internal military use, to prevent adversaries from reverse-engineering the Space Surveillance Network's sensor capabilities breakingdefense.com.
The commercial implication is precise. While some commercial entities have been given limited access to covariance data in the past, no other firm besides SpaceX has been given routine access to those error bubbles for non-government uses breakingdefense.com. Translated into plain commercial terms: SpaceX can now run its Stargaze collision-avoidance service on government-grade positional intelligence that its nearest competitors cannot see. There is a restriction attached: the JCO-SpaceX agreement does contain certain stipulations, and SpaceX is "not allowed to use that for any sort of commercial gain or benefit" breakingdefense.com. Whether that restriction proves durable or enforceable in practice is a separate question from whether it confers a structural information advantage in the interim.
What SpaceX Is Trading
To understand why the JCO agreed to this, you need to understand what Stargaze actually is.
Stargaze uses data collected from nearly 30,000 star trackers, each of which makes continuous observations of nearby objects, resulting in approximately 30 million transits detected daily across the fleet breakingdefense.com. That figure, 30,000 trackers, refers to individual sensor units distributed across the constellation, roughly three cameras per spacecraft on approximately 10,000 satellites 
newatlas.com. Stargaze utilises existing hardware on the Starlink constellation to track thousands of objects in near real-time, offering a significant upgrade over traditional ground-based radar systems 
keeptrack.space. Traditional space situational awareness relies heavily on ground-based sensors that observe satellites only a few times per day, often leading to large uncertainties in orbital predictions keeptrack.space.
The scale of that constellation has grown substantially since Stargaze launched. The KeepTrack Space Brief for September 2026 records approximately 11,133 Starlink satellites in orbit in total keeptrack.space. When SpaceNews reported on Stargaze at its January 2026 unveiling, it described a system using data from star tracker cameras on Starlink satellites to track satellites and debris in low Earth orbit, with SpaceX stating that nearly 10,000 Starlink satellites in orbit can provide 30 million observations of objects each day 
spacenews.com. The constellation has continued to grow since.
SpaceX opened Stargaze to external satellite operators in August 2026, offering conjunction screening at no charge 
satnews.com. Under the classified data-sharing arrangement, SpaceX provides Stargaze sensor data to the JCO and in return can use HAC covariance data to refine its own collision assessments breakingdefense.com. The operational stakes are real. Starlink satellites fired their thrusters to dodge a collision roughly 300,000 times in 2025, about 822 times a day keeptrack.space. That 2025 calendar-year figure has since been surpassed: Starlink satellites performed 207,152 avoidance maneuvers between December 2025 and May 2026 alone, bringing the yearly total from June 2025 to May 2026 to over 355,000, more than three times as many as the constellation performed in 2024 
space.com.
The Space Force has been building toward commercial integration for some time. Within the Space Force, the JCO is the main organisation that buys data and services, and it expended $76.8 million on data and services from the Global Data Marketplace from January 2023 through September 2025, with purchases supporting missions including space domain awareness and tactical surveillance, reconnaissance and tracking 
gao.gov. That $76.8 million represents the JCO's share of a larger Space Force-wide total: from January 2023 through September 2025, the Space Force purchased $123.9 million of commercial data and services through the Global Data Marketplace in total, with JCO accounting for $76.8 million (62%) and System Delta 810 accounting for $35.6 million (29%) of that spending gao.gov. The SpaceX arrangement is the sharpest expression of that integration policy so far, but it sits within a broader architectural shift toward treating commercial sensing as an operational input.
Why This Is a Problem for Independent SDA Ventures
Stargaze, even if its free tracking service is based only on the observations of its own cameras, is at a minimum a market perturbator satnews.com. With classified covariance data layered on top, it becomes something more: a vertically integrated service combining sensor scale no competitor can replicate, government-grade positional accuracy previously unavailable commercially, and a free pricing model subsidised by the Starlink revenue base.
The commercial SDA market that independent firms are competing for is real but modest, and figures from different research firms vary enough to warrant careful reading. Fortune Business Insights projects the global SSA market will grow from USD 2.33 billion in 2026 to USD 5.04 billion by 2034, at a CAGR of 11.02% 
fortunebusinessinsights.com. MarketsandMarkets, using a different methodology, projects the SSA market at USD 1.73 billion in 2025 growing to USD 2.79 billion by 2030 at a CAGR of 10.0% 
marketsandmarkets.com. Treat the range as the signal; no single estimate should be cited as authoritative.
Within that market, the European angle matters and is quantifiable. According to Coherent Market Insights, Europe is expected to be the second-largest SSA market globally, accounting for over 32.2% of market share in 2026, a position driven by ESA initiatives and rising national space budgets across the continent 
coherentmarketinsights.com. European commercial SDA firms have been growing. Portugal's Neuraspace, founded in 2020 and providing AI-enabled space domain awareness and space traffic management solutions to commercial, institutional, and governmental customers worldwide, closed a EUR 15.6 million round in August 2026 
eu-startups.com.
The SpaceX data deal does not foreclose that market. But it permanently changes the precision benchmark. Any European SDA venture pitching on conjunction accuracy now needs to explain how its uncertainty budgets compare to a system that feeds on classified US covariance data. That is not a comfortable position in a procurement conversation.
The Structural Risk: Data as Moat
The Department of Defense relies on companies in the commercial space sector to provide data about activities in space, but a GAO review found that some officials are missing opportunities to use this data due to licensing costs, perceived use restrictions, and concerns over long-term data access gao.gov. The JCO's broader model is to leverage diverse commercial capabilities for non-classified SDA services 
sscfrontdoor.experience.crmforce.mil. The SpaceX arrangement diverges from that model in a specific and important direction: it is exclusive, involves classified data flowing to a private operator, and has no reported open pathway for other firms to access equivalent terms breakingdefense.com.
The asymmetry this creates is the real risk for independent ventures. No other firm besides SpaceX has been given routine access to the HAC covariance error bubbles for non-government uses breakingdefense.com. The GAO's August 2026 review documents that commercial sensing and analytics have moved well beyond isolated demonstrations, with the Space Force purchasing $123.9 million of commercial data and services through the Global Data Marketplace from January 2023 through September 2025 gao.gov. Most JCO spending went to electro-optical and radar observation data purchased from multiple vendors gao.gov. The SpaceX arrangement is structurally different: it is not a purchase of SpaceX data at a market rate, but a classified barter that produces a durable, asymmetric information advantage for a single firm.
For Founders
If you are building in SDA or space traffic management: The SpaceX data advantage is real and likely durable. Do not compete on raw positional accuracy against a system with classified covariances and approximately 30,000 on-orbit sensors producing 30 million observations a day 2 sources. Compete instead on legal addressability in European and allied markets where US classified data cannot be offered commercially, on transparency of uncertainty propagation for regulatory licensing purposes, and on dual-use intelligence applications that European governments will pay premium prices to keep in sovereign hands.
If you are building sensors or analytics adjacent to SDA: The JCO's broader ambition is to integrate more commercial data into military operations, not just SpaceX's. The fiscal year 2026 defence funding agreement provided an additional $30 million for Tactical Surveillance, Reconnaissance and Tracking and $49.5 million to examine commercial services involving positioning, navigation, timing, environmental monitoring, and space domain awareness requirements 
newspaceeconomy.ca. Speed and breadth of coverage, not just accuracy, are active procurement criteria. Stargaze currently uses about 30,000 star-tracking cameras mounted on Starlink satellites to track objects in low Earth orbit, but Starlink satellites are "bad at seeing the small stuff," meaning radar data remains an extremely valuable supplementary tool breakingdefense.com. European sensor networks covering orbital regimes where Stargaze has lower coverage, particularly GEO and highly elliptical orbits rarely overflown by LEO constellations, remain differentiated plays.
On the data-moat principle broadly: This deal is a clean example of a pattern that will repeat across dual-use space and defence markets. When a dominant infrastructure operator offers a free service and then converts government data access into a structural cost advantage, the independent market does not disappear but it stratifies. The upper tier, precision services for government customers with the highest accuracy requirements, goes to whoever holds the classified data relationship. The lower tier, commercial operators wanting basic conjunction screening, gets commoditised. European ventures need to be deliberate about which tier they are building for and price their runway accordingly. The Neuraspace EUR 15.6 million round shows capital is still available for the right positioning eu-startups.com, but the position needs to be genuinely defensible from today, not from a pre-Stargaze baseline.
On US market access: A European SDA venture seeking a JCO data-sharing arrangement now faces ITAR constraints, security clearance requirements, and the precedent that the most sensitive data tier has already been allocated to a single firm with no equivalent open pathway breakingdefense.com. Plan US market entry carefully. The JCO spent $76.8 million through the Global Data Marketplace as the main organisation that buys data and services gao.gov, and that market remains open to non-classified sensing from multiple vendors. But do not assume symmetry with what SpaceX has been given, and do not build a business model that depends on eventually accessing HAC covariances on equivalent terms.
&w=3840&q=75)
&w=3840&q=75)
&w=3840&q=75)
&w=3840&q=75)