The short version

• The EU Space Act is a proposed regulation, COM(2025) 335 final, published 25 June 2025. It is not law yet.
• 120 articles across three pillars: safety, resilience and sustainability.
• It reaches non-EU operators. Obligations attach to anyone providing space based data or services in the Union, wherever they are established.
• The Council and Parliament are still negotiating. Application dates have already slipped from the Commission’s proposed 1 January 2030.

Europe currently regulates space activity through thirteen different national regimes. A satellite operator established in Luxembourg, launching from French Guiana, with a ground station in Spain and customers in Germany, deals with several authorisation frameworks that were written independently and do not align.

The EU Space Act is the Commission’s attempt to replace that with one instrument. It is ambitious, it is contested, and as at August 2026 it is still a proposal.

What was proposed

The full title is the Proposal for a Regulation of the European Parliament and of the Council on the safety, resilience and sustainability of space activities in the Union. It was published on 25 June 2025 as COM(2025) 335 final, under procedure 2025/0335(COD), with Article 114 TFEU as its legal basis. That choice of legal basis is deliberate: this is framed as an internal market measure, not a space policy one.

The text runs to 120 articles across seven titles.

Pillar one: safety

Tracking of space objects, debris mitigation, collision avoidance and end-of-life disposal, supported by a Union Register of Space Objects. This is the pillar that most resembles what the FCC and the FAA already do, and where operators with US authorisations will find the least new ground.

Pillar two: resilience

Cybersecurity risk management obligations across the full mission lifecycle, from design through to disposal. This pillar is more detailed than NIS2 and was originally drafted as lex specialis to it, which raised an obvious question about which instrument governs an operator caught by both.

Pillar three: sustainability

Environmental footprint declarations, a space specific life cycle assessment methodology, sustainability performance classification, and design requirements for debris limitation and in-space servicing. Minisatellites are carved out of some of the servicing design requirements.

The extraterritorial reach

This is the provision non-European operators need to read.

Article 2 brings in Union space operators providing services in the Union, third country space operators providing space based data or space services in the Union, international organisations operating space assets, primary providers of space based data, and advanced space service providers such as collision avoidance and in-space operations providers.

The connecting factor is service provision, not establishment. Obligations attach irrespective of where the operator is established, if the data or services are provided in the Union. Safety and collision avoidance rules extend to objects at or below geostationary orbit that generate data or enable services in the Union.

If the final text keeps that structure, a US or Indian constellation selling into Europe takes on European compliance obligations regardless of where its licence sits.

Penalties

Member State fines have to be effective, proportionate and dissuasive, with no stated ceiling in the proposal. Separately the Commission may impose fines of up to twice the profits gained, twice the losses avoided, or 2 percent of worldwide annual turnover, plus daily penalty payments for up to six months.

The 2 percent of global turnover figure is the one that has focused minds. It is GDPR-shaped, and it is being applied to an industry with a much smaller compliance function.

Where the file has got to

The lead committee is ITRE, with Elena Donazzan of the ECR group as rapporteur.

The rapporteur’s draft report was published on 3 March 2026. Its direction is simplification: cutting the maximum authorisation timeline from twelve months to six, and introducing mutual recognition of national authorisations so an operator authorised in one Member State is not re-examined in another.

ENVI and IMCO adopted opinions in May 2026. ENVI pushes for mandatory environmental impact assessment and lifecycle responsibility. IMCO proposes a “space check” of EU legislation and foreign investment oversight.

On the Council side, the Cypriot Presidency circulated a compromise text on 30 March 2026, with all Member States maintaining scrutiny reservations. A progress report on 8 May 2026, ahead of the Competitiveness Council on 29 May, described the text as moving in the right direction and reducing complexity, while listing open issues: regulatory scope and dual use exemptions, governance architecture, avoiding a duplicative Union layer over national procedures, and the equivalence regime for third country operators.

That last one is the sleeper issue for non-EU operators. Council changes so far would keep a registration and e-certificate requirement for third country operators while exempting them from direct Space Act obligations where an equivalence finding applies. Whether that survives is the single most commercially significant open question in the file.

Other Council changes worth tracking

• International organisations removed from direct obligations, with applicability made contingent on an EU to IO agreement
• Collision avoidance service providers excluded from scope
• The NIS2 relationship changed from lex specialis to parallel application, with large operators staying under NIS2 and smaller and third country operators under the Space Act

When would it apply

The Commission proposed 1 January 2030, with 1 January 2032 for certain asset categories. The Council position is 36 months after entry into force, with a transitional period of eight years for specific categories. The Parliament rapporteur is aligned with the Council on timing.

Translated: nobody should be planning compliance spend for 2030. A realistic first application date is later, and the transitional arrangements for existing constellations will matter more than the headline date.

One correction worth making

A number of secondary sources have described COM(2026) 152 of 7 April 2026 as an amended Space Act proposal. It is not. That document is the Proposal for a Regulation on the European Union Space Services Agency, procedure 2026/0084(COD), which gives EUSPA a standalone founding regulation, renames it, and raises its budget from €525.7 million for 2021 to 2027 to €979.6 million for 2028 to 2034. It is a separate instrument.

There is no amended Commission proposal for the Space Act. The text is being changed through Council compromise texts and Parliament committee amendments, which is why tracking it requires reading committee documents rather than waiting for a new COM number.

The short version

• The FCC replaced the 25 year post-mission disposal guideline with a five year maximum in Second Report and Order FCC 22-74, adopted 29 September 2022.
• It is codified at 47 CFR 25.283(e) and became effective 9 September 2024, with compliance required from 29 September 2024.
• It applies to spacecraft ending their mission in or passing through low Earth orbit below 2,000 km and planning uncontrolled reentry, including non-US systems granted US market access.
• Satellites already in orbit are grandfathered. Anything launched after 29 September 2024 is not.

For twenty five years the working number in orbital debris mitigation was twenty five years. A satellite in low Earth orbit was expected to clear the region within a quarter century of the end of its mission. That figure came out of NASA analysis in the 1990s and worked its way into the Inter-Agency Space Debris Coordination Committee guidelines, then into national practice.

It was a reasonable number for a world launching dozens of satellites a year. It is not a reasonable number for a world launching thousands.

What the rule says

The FCC’s Second Report and Order, FCC 22-74, was adopted on 29 September 2022 in IB Docket Nos. 18-313 and 22-271. The core change is short: for the spacecraft it covers, post-mission disposal must be completed as soon as practicable and no more than five years after the end of mission.

The obligation is codified at 47 CFR 25.283(e), with parallel amendments for experimental licences at 47 CFR 5.64 and amateur operations at 47 CFR 97.207.

Three limits on scope matter.

Orbit. It applies to space stations that will terminate their mission in, or pass through, low Earth orbit below 2,000 km, and that plan uncontrolled atmospheric reentry. Geostationary disposal continues to be handled by the graveyard orbit requirement, which is a separate obligation and the one DISH failed, as we discuss in our case note on that consent decree.

Reach. It applies both to systems licensed by the FCC and to non-US systems granted access to the US market. This is the important commercial point. A non-US constellation that wants to sell service in the United States takes on the obligation.

Timing. The order was adopted in September 2022 but the rule became effective on 9 September 2024 with compliance required from 29 September 2024, following the standard Paperwork Reduction Act sequence. Satellites already on orbit are exempt. Systems that were authorised but not yet launched had a two year grandfathering window running from adoption. Anything launched after 29 September 2024 complies or does not fly.

Why five years and not zero

Five years is not a physics number. It is a compromise between two positions.

Operators argued for a longer window on cost grounds. Deorbit capability is mass, mass is money, and for a cubesat the propulsion needed to guarantee a rapid deorbit can be a meaningful fraction of the spacecraft.

Debris researchers argued that even five years leaves a long tail. The relevant risk is not one dead satellite. It is the aggregate population of uncontrolled objects in the shells where constellations operate, and the probability that any two of them meet.

The Commission landed on five years plus a “as soon as practicable” obligation, which does real work. An operator that can deorbit in one year cannot elect to take five.

What it means for an application

For most operators the rule does not change the application form so much as the engineering that has to sit behind it. Three things need to be demonstrable at filing.

Disposal capability

Either propulsive deorbit, a drag device, or an orbit low enough that natural decay does the job inside five years. Natural decay is the cheapest route and it is why so many smallsat constellations cluster below about 550 km. Above roughly 600 km, passive compliance stops being realistic.

Reliability

The FCC expects a stated post-mission disposal reliability. If a percentage of the constellation is going to fail before disposal, that has to be in the analysis, and for large constellations a small failure rate is a large absolute number.

Casualty risk

Uncontrolled reentry raises human casualty risk on the ground. The working threshold used across the US agencies is one in ten thousand for a given reentry event, and design for demise is the usual answer.

Where the FCC has gone since

The five year rule now sits inside a much larger overhaul. In October 2025 the Commission proposed replacing Part 25 wholesale with a new Part 100, and on 22 July 2026 it adopted a Report and Order, FCC 26-47, doing exactly that.

The headline changes for operators are procedural rather than substantive on debris: a uniform 15 day public notice period, a 30 day deadline for the Commission either to put an application on public notice or say what is missing, a 60 day decision deadline after comments close, a stated presumption in favour of granting conforming applications, voluntary annual NGSO processing rounds, $10 million surety bonds for processing round applicants declining to zero at 90 percent deployment, and 20 year licence terms as standard.

Part 25 remains in force until the Space Bureau announces the effective date of Part 100 by public notice. Until then, cite Part 25.

The honest assessment

The five year rule is the most consequential unilateral debris measure any regulator has taken, and it was taken by a communications regulator using its licensing power rather than by a space agency or a treaty body. That is worth sitting with.

It also shows the limits of that route. The FCC can only reach systems it licenses or that want US market access. It cannot reach a constellation that never sells into the United States, and it has no jurisdiction over the objects already up there, which is where the accumulated risk lives.

The short version

• Part 450 has been the only route to a US launch or reentry licence since 10 March 2026, when the legacy Parts 415, 417, 431 and 435 were removed from the CFR.
• It is performance based. One licence can cover multiple vehicle configurations, mission profiles and sites.
• The FAA had issued 14 Part 450 licences as at March 2026.
• Two things changed the economics in 2026: a new per launch user fee, and a July 2026 proposal to waive NEPA and twelve other environmental statutes for launch licensing.

If a rocket leaves the ground from US territory, or a US company launches from anywhere, the FAA Office of Commercial Space Transportation has to have said yes first. The instrument is a vehicle operator licence under 14 CFR Part 450, and since March 2026 there is no alternative route.

That deadline mattered more than most regulatory transitions, because it retired four separate rule parts that some of the largest operators had been flying under for years.

What Part 450 replaced, and why

The Streamlined Launch and Reentry License Requirements rule was published on 10 December 2020 at 85 FR 79566 and took effect on 10 March 2021. It folded four regimes into one:

• Part 415, launch licences
• Part 417, launch safety
• Part 431, reusable launch vehicle mission licences
• Part 435, reentry

The old structure asked a different set of prescriptive questions depending on which box your vehicle fell into. If your vehicle did not fit a box, and by 2018 many did not, you negotiated waivers.

Part 450 replaced prescription with performance. Instead of telling an operator which specific analysis to run, the rule sets a safety outcome and asks the operator to propose a means of compliance. The FAA either accepts it or does not.

The commercial payoff is scope. A single Part 450 licence can authorise multiple vehicle configurations, multiple mission profiles and multiple launch or reentry sites. Under the old parts, a new site or a materially different mission often meant a new application.

The transition deadline that just passed

Operators holding legacy licences were given five years. All licences issued under Parts 415, 417, 431 and 435 expired by 9 March 2026, and the legacy parts came out of the Code of Federal Regulations on 10 March 2026.

Operators who completed the move include Blue Origin for New Shepard, Firefly for Alpha, SpaceX for Falcon 9, Falcon Heavy and Dragon, Rocket Lab for Electron, and ULA for Atlas and Vulcan. The FAA said in March 2026 that it had issued 14 Part 450 licences since the rule took effect.

Anyone still assuming a legacy path exists is working from stale material.

How the application is structured

Part 450 runs in five subparts, and it helps to think of them as stages rather than chapters.

Subpart C is where the work is. Section 450.101 sets the quantitative public safety criteria, expressed as collective and individual risk thresholds for the public, aircraft and ships. Everything downstream, the flight safety analysis, the hazard control strategies, the debris analysis, exists to demonstrate that those numbers are met.

The practical friction point is the means of compliance process. An operator proposes a method, the FAA accepts it, and only then can the analysis built on it be relied on. Applicants who leave means of compliance to the end of the drafting process lose months.

Financial responsibility runs in parallel

A Part 450 licence does not stand alone. Part 440 sits alongside it and sets the insurance obligation.

The FAA determines maximum probable loss, the greatest dollar amount of loss for bodily injury or property damage reasonably expected from the licensed activity, and issues a determination within 90 days of a complete request. The probability thresholds are one in ten million for third parties and one in one hundred thousand for government personnel and property.

Statutory caps sit above that: $500 million for third-party liability insurance and $100 million for government property, or the maximum available on the world market at reasonable cost if that is lower. Above the insured layer, 51 U.S.C. 50915 provides for government indemnification of up to $1.5 billion indexed from 1989, subject to appropriation that has never been made.

We walk through how that stack interacts with the contract in our piece on cross waivers and indemnification.

What changed in 2026

Launch user fees

The One Big Beautiful Bill Act, Pub. L. 119-21, added 51 U.S.C. 50924 on 4 July 2025. From 2026 there is a per launch or reentry fee, calculated as the lesser of a per pound of payload rate or an annual cap. For 2026 that is $0.25 per pound with a $30,000 cap, rising annually to $1.50 and $200,000 by 2033, then indexed to CPI.

The FAA’s implementing notice was published on 22 April 2026 under Docket FAA-2026-4170. Operators must supply payload weight at least 60 days before a mission and pay within 30 days of fee notification. For a heavy lift operator the cap makes this rounding error money. For a small launch operator flying frequently it is a line item worth modelling.

The environmental waiver proposal

The bigger story is procedural. On 30 July 2026 the FAA published an NPRM under Docket FAA-2026-8614, RIN 2120-AM51, that would insert a new section 400.3 disapplying thirteen environmental and resource protection statutes to launch and reentry licensing.

The list is not marginal. It includes NEPA, the Endangered Species Act, the Clean Water Act, the Clean Air Act, the Coastal Zone Management Act, the National Historic Preservation Act, the Marine Mammal Protection Act and the Magnuson-Stevens Act. The claimed authority is 51 U.S.C. 50905(b)(2)(C). Comments closed on 31 August 2026.

Environmental review has been the main source of delay and the main litigation surface for launch site expansion. If this rule is finalised in anything like its proposed form, that surface largely disappears, and the practical constraint on launch cadence shifts back to range availability and airspace.

It also runs in the same direction as the FCC’s proposal to treat space operations as extraterritorial and outside NEPA entirely, which we cover in our case note on the Dark-Sky litigation.

The Part 450 rewrite itself

An Aerospace Rulemaking Committee was chartered in November 2024 to recommend improvements to Part 450 across nine areas, including reentry operations, early test flights, means of compliance approval and the definition of safety critical systems. An executive order signed on 13 August 2025 directed the Secretary of Transportation to reevaluate and amend or rescind Part 450.

As at August 2026 no committee report and no amendment NPRM had appeared on the FAA’s rulemaking docket. Treat the substantive rewrite as still pre-proposal.

The learning period, correctly stated

Commentary routinely describes a “moratorium” on human spaceflight safety regulation. The statute is narrower than that.

Under 51 U.S.C. 50905(c)(9), beginning on 1 January 2028 the Secretary may propose occupant safety regulations without regard to the statutory restrictions in subparagraphs (C) and (D). The current date was set by Pub. L. 118-159 on 23 December 2024, the latest in a series of extensions running back to the 2004 Commercial Space Launch Amendments Act.

It restricts proposing regulations, and it is subject to triggers including a serious injury, fatality or high risk design event. It is not a blanket prohibition on the FAA acting.