UK reveals £29.5m military SpaceX spending

The UK Ministry of Defence holds approximately 1,000 Starshield terminals and 500 Starlink terminals, putting numbers around the growing military use of SpaceX low-Earth-orbit communications services.

Figures released through freedom-of-information disclosures show expenditure of around £13 million on Starshield and £16.5 million on Starlink, covering terminals and communications services. The combined £29.5 million spend remains small beside major sovereign satellite programmes, but the equipment quantities show that commercial LEO connectivity has progressed well beyond limited technical trials.

Starlink and Starshield use related SpaceX infrastructure but serve different markets. Starshield is directed towards government and national-security customers and adds service arrangements and security functions intended for military users, while drawing on the scale of the company’s wider satellite network.

The UK began using Starlink during 2022 and subsequently introduced Starshield services, initially through existing equipment before buying dedicated Starshield terminals from 2025. The resulting estate now gives British forces access to a relatively large pool of deployable commercial satellite terminals alongside established military communications systems.

Commercial capacity joins sovereign satcom

The equipment does not replace Britain’s SKYNET programme. SKYNET remains the sovereign military satellite-communications backbone, with the continuing SKYNET 6 programme intended to provide new satellites, ground infrastructure, user equipment, and service capability into the 2040s.

The two approaches offer different characteristics. Sovereign military satellites provide national control over strategically important infrastructure, while commercial LEO networks can offer large numbers of spacecraft, comparatively low latency, wide coverage, and terminals that can be deployed quickly without waiting for a dedicated military satellite programme.

Operating both creates a more layered communications architecture, but it also introduces additional integration work. Military networks have to determine which traffic can use commercial infrastructure, how encryption boundaries are managed, which terminals are authorised for particular services, and how connectivity moves between sovereign, allied, and commercial systems when availability changes.

Terminal numbers are therefore more revealing than satellite counts alone. Approximately 1,500 pieces of user equipment have to be configured, tracked, patched, powered, transported, and supported, while deployed units need trained personnel who can bring them into service without weakening the wider network’s security posture.

SpaceX’s infrastructure also remains externally owned. The MoD can procure large quantities of terminals and airtime without controlling the spacecraft constellation, ground network, core software, launch programme, or commercial decisions behind the service, creating dependencies different from those associated with nationally owned satellites.

Starshield is designed to address part of that gap by providing a service tailored to security customers. Features associated with the offering include enhanced encryption and priority arrangements, while dedicated gateways and terminals allow military users to operate within a more controlled service environment than ordinary consumer broadband.

Those additional requirements affect cost. Reuters reported that Starshield service packages available to the MoD are considerably more expensive than normal Starlink subscriptions, reflecting the different service conditions, data allowances, prioritisation, and security arrangements attached to government use.

The comparison cannot be reduced to the monthly price of connectivity. A military service has to remain usable during congestion, disruption, or crisis conditions, and the value of priority access or government-specific support only becomes apparent when ordinary commercial demand is competing for the same underlying infrastructure.

Britain’s approach increasingly reflects an assumption that resilience will come from combining networks rather than relying on one orbital system. SKYNET can provide sovereign assured capacity, while commercial services offer another path for traffic and rapid access to a large distributed constellation.

That architecture still has common points of failure. Ground stations, terrestrial backhaul, network-management systems, electrical power, cyber controls, and user terminals all remain vulnerable irrespective of how many satellites are overhead, while interoperability between separate networks has to be engineered and exercised before it is needed under pressure.

The MoD’s disclosed terminal estate makes those integration questions increasingly operational rather than theoretical. Thousands of users can make use of commercial connectivity only if identity, routing, security policy, configuration control, and support systems scale alongside the hardware.

Britain is continuing sovereign investment at the same time. SKYNET 6A is being manufactured by Airbus Defence and Space UK as the first new satellite element in the programme, while wider SKYNET 6 work is intended to modernise ground and user infrastructure as well as the space segment.

Commercial services can therefore supplement a national programme without removing the requirement for sovereign engineering and manufacturing capability. The harder problem is ensuring that the layers complement one another, rather than creating separate terminal fleets and networks that cannot exchange traffic cleanly during disruption.

The expenditure disclosed so far is modest by strategic communications standards, but the approximately 1,500 terminals make the scale difficult to dismiss as experimentation. SpaceX connectivity is already becoming part of the equipment available to British defence users.

Future procurement will show how permanent that role becomes. Spending levels, terminal growth, integration with SKYNET 6, and the adoption of alternative commercial providers will indicate whether Starshield becomes one component in a deliberately diversified architecture or develops into a dependency that becomes increasingly expensive to unwind.