The short version

• Nobody can own the Moon. Article II of the Outer Space Treaty is unambiguous about the place itself.
• Four states have legislated that companies can own what they extract: the United States in 2015, Luxembourg in 2017, the UAE in 2019 and Japan in 2021.
• Seventy states have now signed the Artemis Accords, whose Section 10 records the view that extraction is not by itself national appropriation.
• The COPUOS working group on space resources is due to prepare draft principles during 2026, with adoption targeted for 2027.

The question gets asked at every space conference and it is almost always asked badly. “Who owns the Moon” has a clean answer: nobody, and nobody can. The interesting question is narrower. If a company lands a machine on the Moon and it fills a hopper with regolith, who owns the regolith?

That question is genuinely unsettled, and the way it gets settled will determine whether lunar and asteroid mining is financeable.

What Article II settles

Outer space, including the Moon and other celestial bodies, is not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.

Outer Space Treaty, Article II

That rules out territory. No flag plant creates a claim, no continuous occupation ripens into title, and the closing phrase “by any other means” was drafted to catch creative attempts.

Article VI then closes the obvious loophole. States are internationally responsible for the activities of their non-governmental entities in space, so a company cannot do what its state cannot do. The idea that a private company could claim lunar territory where a state could not has no support in the text.

Everyone agrees on all of that. The disagreement starts one step later.

The consistency argument

The case that extraction is lawful runs like this.

Article II prohibits appropriation of the celestial body. Regolith removed from the surface is no longer the celestial body, in the same way that a fish taken from the high seas is no longer the high seas. Under UNCLOS Part VII no state may appropriate the ocean, and yet harvested fish are unambiguously owned. The analogy is not perfect but it is close, and it reflects a very old distinction between a commons and the resources taken from it.

Article I independently protects “use” of outer space, and extraction is a form of use. The drafters chose broad language and did not carve out resource activity.

Finally, Article II’s list of prohibited means, sovereignty, use, occupation and any other means, is aimed at sovereign claims. On this reading, Section 10 of the Artemis Accords does not change the law. It states it.

The inconsistency argument

The case against is not merely a policy objection.

Article VI attributes non-governmental activity to the licensing state. If a state authorises and continuously supervises a mining operation, and confers property rights in the output, the state is doing through a licensee what Article II forbids it to do directly.

Commercial scale extraction is not a fishing boat passing through. It requires a fixed installation, exclusive access to a site with the right illumination and volatile content, and that exclusivity has to be defensible over decades. The Accords make this explicit through the Section 11 concept of safety zones, which critics read as territorial exclusion under another name. There are only a handful of permanently shadowed craters at the lunar south pole with useful water ice, and the first operator to occupy one has, in practical terms, taken it.

There is also a process objection. A group of states is generating interpretive practice through a non-treaty instrument, outside COPUOS, where the consensus rule gives every state a voice. Russia and China have not signed and have objected publicly to the approach.

What the national laws actually say

The US provision, at 51 U.S.C. 51303, entitles a US citizen engaged in commercial recovery of an asteroid or space resource to possess, own, transport, use and sell what they obtain, in accordance with applicable law including US international obligations. Section 403 of the same Act adds that Congress does not thereby assert sovereignty or exclusive rights over any celestial body.

That disclaimer is doing careful work. The statute is drafted to confer a right against other US persons in US courts, not to assert a claim erga omnes.

Luxembourg’s Article 1 is the shortest and most quoted provision in the field: space resources are capable of being owned. The rest of the law is an authorisation regime, requiring prior written mission authorisation, a Luxembourg-law company or an EU company with a Luxembourg branch, and disclosure of shareholders and management.

Japan’s version is the most procedurally integrated. A permit is granted alongside a satellite launch authorisation, the applicant files a business activity plan, the Prime Minister decides in consultation with METI, and approved permits and plans are published. The person holding the permit owns the resources exploited in accordance with the approved plan.

Where the Accords have got to

The Artemis Accords opened for signature on 13 October 2020 with eight founding signatories. As at August 2026 there are 70, with Serbia signing on 16 July 2026 and Mauritius on 17 July 2026. Eleven states signed during 2026 alone.

The relevant text is Section 10, which notes that extraction of space resources does not inherently constitute national appropriation under Article II, and commits signatories to conduct extraction consistently with the Outer Space Treaty and to inform the UN Secretary-General, the public and the international scientific community of their extraction activities.

The Accords are a political commitment, not a treaty. They create no new binding obligations and are not registered under Article 102 of the UN Charter. What seventy signatures do create is state practice, and state practice is one of the ingredients of customary international law. That is precisely why the number matters and why the states outside it have objected.

The multilateral track nobody talks about

The COPUOS Legal Subcommittee established a Working Group on Legal Aspects of Space Resource Activities in 2021, with a five year workplan running from 2022 to 2027.

Information gathering and legal analysis ran from 2022 to 2024, including an international expert conference in Luxembourg on 26 March 2024. Views on preliminary recommendations were exchanged during 2025. Draft principles are to be prepared during 2026, with finalisation and adoption by the Committee targeted for 2027. The most recent circulated draft of the initial set of recommended principles is dated 17 December 2025.

This is the counterweight story, and it gets a fraction of the coverage the Accords do. If COPUOS produces a consensus set of principles in 2027, the Accords will look like an early expression of a broader position. If it does not, the Accords will be the only text in the field, and the divide between signatories and non-signatories becomes the governing structure by default.

What a financier should take from this

For anyone underwriting a lunar or asteroid venture, the honest position is that ownership of extracted material is well supported in the domestic law of the four legislating states and unsettled in international law.

That is not necessarily fatal. Most of the disputes that would actually arise, over offtake, over trespass on a site, over financing security, would be litigated between private parties in national courts applying national law. The international question bites at the margin: if a non-signatory state disputes the operation diplomatically, or if the operator needs recognition of its title in a jurisdiction that rejects the position.

The gap that has no answer at all is priority. There is no international registry of claimed extraction sites, no first-in-time rule, and no forum to resolve two operators wanting the same crater. Safety zones under the Accords are a notification and consultation mechanism among signatories, not a title system.

The short version

• The Outer Space Treaty is about 2,500 words long, entered into force on 10 October 1967, and had 118 states parties as at 1 January 2026.
• Six articles do almost all the work: I, II, VI, VII, VIII and IX.
• Article VI is the commercial one. It makes states answerable for their companies, which is why licensing regimes exist.
• The treaty settles almost nothing about resources, debris, traffic or brightness. That is not a drafting failure. Those problems did not exist in 1966.

Most explanations of the Outer Space Treaty start with the phrase “province of all mankind” and end with a vague gesture at peaceful use. That is the ceremonial reading. The working reading is narrower and more useful, because six of its seventeen articles carry essentially all the legal weight.

The treaty was adopted by UN General Assembly resolution 2222 (XXI) on 19 December 1966, opened for signature in London, Moscow and Washington on 27 January 1967, and entered into force on 10 October 1967. It was negotiated in about eighteen months by two superpowers who both wanted to stop the other one claiming the Moon.

Here is what each of the load-bearing articles actually does.

Article I: freedom of use, for everyone

Exploration and use of outer space shall be carried out for the benefit and in the interests of all countries, and outer space shall be free for exploration and use by all states without discrimination.

Two practical consequences. First, there is no permission requirement in international law to go to space. You do not apply to anyone. Second, “use” is a protected activity and it is broad. The argument that extracting resources is a form of “use” protected by Article I starts here.

The “province of all mankind” phrase sits in this article. It is aspirational language, not a property rule. It has never been read as creating a common heritage regime, which is precisely what the Moon Agreement tried to add twelve years later and why so few states joined it.

Article II: the non-appropriation rule

Outer space, including the Moon and other celestial bodies, is not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.

Outer Space Treaty, Article II

One sentence, and the most litigated sentence in the field despite never having been litigated.

Note what it covers. It prohibits appropriation of the place. It says nothing explicit about material removed from that place. The drafters were worried about territorial claims of the kind that carved up Antarctica and Africa. They were not thinking about regolith processing.

That silence is the crack that four national space resources laws and Section 10 of the Artemis Accords have grown into. We set out both sides of that argument in our piece on space resources.

The other thing to notice is the phrase “by any other means”. Critics of the resources position lean on it hard. If a company operates an exclusive mining site for twenty years under a national licence, they argue, that is occupation dressed up in commercial clothing, and the Accords’ safety zones concept makes the exclusivity explicit.

Article VI: the article that pays your fees

States parties bear international responsibility for national activities in outer space, whether carried on by governmental agencies or by non-governmental entities, and the activities of non-governmental entities require authorisation and continuing supervision by the appropriate state party.

This is the hinge between international law and commercial practice.

A private company is not a party to the treaty and cannot be. So the treaty makes the state responsible for private conduct, and then requires the state to authorise and supervise it. Every national licensing regime in existence is a state discharging that obligation.

Note “continuing supervision”. The obligation does not end at grant. It is why licences carry ongoing conditions, reporting duties and end-of-life requirements, and why the FCC could treat DISH’s disposal plan as an enforceable licence condition years after the satellite launched.

Article VII: liability, in outline

Each state party that launches or procures the launching of an object into outer space, and each state party from whose territory or facility an object is launched, is internationally liable for damage caused by that object.

Four categories of launching state, and they can all be liable for the same object. A satellite built in Germany, procured by a Luxembourg company, launched from French Guiana on a US vehicle can implicate several states at once.

Article VII is the outline. The 1972 Liability Convention fills in the detail, splitting damage into absolute liability on the surface of the Earth and fault liability everywhere else.

Article VIII: jurisdiction follows registration

The state of registry retains jurisdiction and control over the object and any personnel on it, and ownership of objects is not affected by their presence in outer space or by their return to Earth.

This does two useful things. It supplies the connecting factor for criminal and civil jurisdiction on board, which is how the Intergovernmental Agreement on the Space Station allocates jurisdiction module by module. And it confirms that a satellite remains the property of its owner while in orbit, which is the foundation for satellite financing, security interests and salvage arguments.

The 1975 Registration Convention builds the machinery: launching states file object data with the UN Secretary-General. It has 77 parties, and compliance is patchy for smallsats.

Article IX: due regard, contamination and consultation

States shall conduct activities with due regard to the corresponding interests of other states, shall avoid harmful contamination of space and adverse changes in the Earth’s environment, and shall consult before proceeding with an activity that would cause potentially harmful interference.

Article IX is the closest the treaty comes to an environmental provision, and it is doing a lot of unglamorous work in modern debates. Debris mitigation guidelines, planetary protection policy and the dark and quiet skies argument all cite it.

Its weakness is that “due regard” and “potentially harmful interference” are undefined, and the consultation mechanism has never been formally triggered by one state against another over a constellation.

What the treaty does not do

It is worth being blunt about the gaps, because a lot of commentary pretends they are not there.

None of this makes the treaty a failure. It has held for nearly sixty years across a Cold War, the collapse of one of its two authors and the arrival of an industry its drafters could not have imagined. But it is a framework, and frameworks need floors built under them. Most of the interesting legal work in space today is that floor building, happening in national rulemaking and in COPUOS working groups rather than in treaty negotiation.

The short version

• Space law is a small stack of UN treaties from 1967 to 1979, sitting on top of a much larger and faster moving pile of national licensing rules.
• The Outer Space Treaty makes states responsible for what their companies do in orbit. That single sentence is why every launch and every satellite needs a government licence.
• Nobody can own the Moon. Whether anyone can own what they dig out of it is the live argument.
• In practice, a commercial space lawyer spends far more time on FAA, FCC, ITU and export control paperwork than on treaty interpretation.

Ask ten people what space law is and most will describe something like maritime law for rockets. That is not far off, but it misses the part that matters commercially. Space law is not one rulebook. It is two layers that behave very differently.

The top layer is international. Five UN treaties negotiated between 1967 and 1979, plus a set of General Assembly principles and guidelines that are not binding but get followed anyway. This layer moves slowly. The last treaty was opened for signature in 1979 and almost nobody joined it.

The bottom layer is national. Every country that lets private operators launch or run satellites has built its own licensing regime, and those regimes are rewritten every few years. This is the layer that decides whether a company can fly.

If you only read the treaties, you will understand the constitutional theory and none of the practice.

The five treaties, and why only three matter day to day

The UN Office for Outer Space Affairs publishes the ratification tables every year. As at 1 January 2026 they looked like this.

Look at the last row. Seventeen parties, none of them a major spacefaring state. The Moon Agreement is the treaty everyone cites in arguments about lunar mining and the one nobody with a rocket has signed. It is a useful piece of evidence about what states were willing to accept in 1979, and very little else.

The Outer Space Treaty is the one that does the structural work. The Liability Convention and the Registration Convention fill in the detail on damage claims and on who has to tell the UN what they launched.

The sentence that built the whole licensing industry

Article VI of the Outer Space Treaty says that states bear international responsibility for national activities in outer space, including activities carried on by non-governmental entities, and that those activities require authorisation and continuing supervision by the appropriate state.

Activities of non-governmental entities in outer space, including the Moon and other celestial bodies, shall require authorization and continuing supervision by the appropriate State Party to the Treaty.

Outer Space Treaty, Article VI

Read that again with a commercial eye. A private company cannot be a party to the treaty. So the treaty reaches private activity by making the state answerable for it. If a US company puts a satellite in orbit and it causes damage, the claim under international law runs against the United States, not the company.

No government accepts that exposure without control. That is why the FAA licenses launches, the FCC licenses spectrum and orbital debris plans, the UK Civil Aviation Authority licenses UK operators, and India routes everything through IN-SPACe. Article VI is the reason those agencies exist in their current form.

Liability: two regimes, one of which has never been used

The Liability Convention splits damage into two categories, and the split matters enormously.

Under Article II, a launching state is absolutely liable for damage caused by its space object on the surface of the Earth or to aircraft in flight. No fault needs to be shown. If a stage falls on a house, the launching state pays.

Under Article III, damage caused anywhere other than the surface of the Earth, which in practice means collisions in orbit, is governed by fault liability. The claimant has to prove the other state was at fault.

The Convention then does something remarkable. It never says what the standard of care is. There is no definition of fault, no orbital rules of the road, no priority scheme. That gap is the single biggest reason the Liability Convention has never been formally invoked for an on-orbit collision, including after the 2009 Iridium and Cosmos collision that produced thousands of trackable fragments.

We wrote about how that plays out in practice in our piece on debris liability.

Nobody owns the Moon. Resources are a different question.

Article II of the Outer Space Treaty is one sentence long and it is the most argued over sentence in the field. Outer space, including the Moon and other celestial bodies, is not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.

That settles territory. It does not obviously settle extracted material. Four states have now legislated to say that a company can own what it extracts: the United States in 2015, Luxembourg in 2017, the UAE in 2019 and Japan in 2021. Seventy states have signed the Artemis Accords, which record the view that extraction does not by itself amount to national appropriation.

Critics answer that Article VI attributes private conduct to the state, so a state licensing extraction is doing indirectly what Article II forbids directly. That argument is not going away, and the COPUOS working group on space resources is due to produce draft principles during 2026.

What a space lawyer actually does

Here is the honest version of the job, for anyone deciding whether to go into it.

Licensing and regulatory

Getting a client through the FAA’s Part 450 launch process, an FCC space station application, an ITU filing through a national administration, or the equivalent in the UK, India, Japan, New Zealand or Australia. This is the bulk of the work and it is closer to aviation and telecoms regulatory practice than to public international law.

Contracts

Launch services agreements, satellite procurement, hosted payload deals, ground segment, data licensing, in-orbit servicing. The distinctive feature is the cross waiver of liability, a reciprocal no-fault waiver that turns the whole supply chain into a set of self-insured silos. We explain how that works in our piece on launch contracts.

Export control

Spacecraft and their components sit on the US Munitions List under ITAR Category XV or on the Commerce Control List in the 9×515 series. Getting this wrong is a criminal exposure, not a commercial one. It occupies a lot of junior time.

Finance and insurance

Launch and in-orbit insurance placement, export credit, satellite backed lending, and increasingly the diligence around megaconstellation financing.

Public international law

Real but small. Treaty interpretation, COPUOS delegation work, advising governments on national legislation, academic writing. If this is the part you find interesting, plan to combine it with something that pays.

Where to start reading

Start with the Outer Space Treaty itself. It is about 2,500 words and you can read it in fifteen minutes. Then read Part 450 or the FCC’s Part 25, not because you will enjoy it, but because it will show you the distance between the constitutional layer and the working layer.

Our directory of regulators and primary sources links straight to the official text of everything mentioned here.