From Blueprint to Build Contract
On August 7, 2026, the European Commission and the SpaceRISE consortium — Eutelsat, SES, and Hispasat — signed the implementation agreement for IRIS² (Infrastructure for Resilience, Interconnectivity and Security by Satellite), closing negotiations that began in January 2026. The deal expands the constellation to 348 spacecraft: 330 in low Earth orbit and 18 in medium Earth orbit, including a reinforced 66-satellite defense and security layer. The Commission says this will lift secure governmental communications capacity by 60% inside the EU and 54% worldwide, with first launches targeted for 2029 (European Commission; EUSPA).
This is a genuine milestone. IRIS² has spent years as a slide deck; it is now a funded, contracted build with named prime contractors, including Hispasat leading ground infrastructure. The total programme cost now stands at roughly €15.6 billion (~$18 billion), funded through the 2028–2034 EU budget cycle plus direct member-state contributions — Poland has committed €656 million, Hungary €500 million, and Spain between €1.6 and €2 billion (Breaking Defense).
The Case for Sovereign Capacity Is Real
The strongest argument for IRIS² is not abstract. On August 2, 2026, President Zelenskyy asked President Trump to press Elon Musk to let Ukraine use Starlink to guide strikes against launch sites inside Russian territory — a request Musk has so far declined, restricting Starlink to Ukrainian and occupied territory only (Fortune). Whatever one thinks of the merits of that specific restriction, the episode is a clean illustration of a structural fact: when a government's battlefield or emergency connectivity runs through a single private foreign operator, that government is negotiating for access rather than commanding it. European defense planners have made this argument for years; this incident gives it a concrete, recent referent rather than a hypothetical one. A continent that wants an independent foreign and defense policy has a legitimate interest in not routing that policy through Elon Musk's discretion.
But the Trade-Off Is Getting Worse, Not Better
The honest accounting, though, is that IRIS² is buying sovereignty at an escalating and uncertain price. The programme's own numbers tell the story: the original 2022 estimate was roughly €6 billion; the December 2024 concession contract came in at €10.6 billion; the August 2026 implementation agreement puts the figure at €15.6 billion — a more than doubling in under four years, entirely before a single operational satellite has flown (Quilty Space). The in-service date has slipped in parallel: 2024, then 2026, then 2028, and now 2029 for first launches with full service later still.
Capability is the harder problem than cost. Independent analysis has pegged IRIS²'s LEO throughput at roughly 2 terabits per second — comparable to two Starlink V3 satellites, but spread across a constellation costing many multiples more. Even Eutelsat's own chief executive has said publicly that IRIS² must compete on price and performance with Starlink and Amazon's Kuiper to win commercial customers, not merely rely on political mandate. That is the correct standard, and it is not yet clearly met.
Steelmanning the Regulator, Then Pushing Back
A fair defender of the programme would say cost comparisons to Starlink are unfair on their face: IRIS² is not trying to be a consumer broadband competitor at Starlink's 7,000-satellite scale. It is purpose-built for a narrower governmental and defense mission where guaranteed availability, EU jurisdictional control, and encryption assurance matter more than raw throughput per euro. On that framing, a resilience-focused sovereign layer is closer in kind to a military communications satellite program than to a retail ISP, and those have always cost more per bit than commercial infrastructure. That argument has real force, and it is why this piece treats sovereign governmental connectivity as a legitimate policy goal rather than mere industrial-policy dressing.
But it does not fully answer the doubling-cost, repeatedly-slipping-schedule pattern, which predates any Starlink-specific incident and looks like ordinary multinational-consortium program management risk rather than a necessary cost of sovereignty. Nor does it resolve a harder question the Commission has so far avoided answering explicitly in its public materials: whether EU institutions and member states will eventually be required, rather than merely encouraged, to route sensitive governmental traffic through IRIS² once it is operational, foreclosing competitive bids from Starlink, Kuiper, or future European commercial entrants like a scaled-up Eutelsat OneWeb constellation.
The Proportionate Path
A sovereignty capability with genuine defense value does not require exclusivity to be justified — it requires that IRIS² earn government business by being reliable and reasonably priced, with competitive commercial alternatives kept available as a check on cost and performance drift. The Commission's own framing — sovereign, secure, and resilient — is compatible with an open governmental-procurement market, not just a single state-favored consortium. If IRIS² cannot ultimately compete on those terms even for defense-sensitive traffic, that is information the EU needs, not a problem to be regulated away by mandate. The €15.6 billion bet is defensible; a mandated monopoly built on top of it would not be.