The announcement that Russia, in partnership with China, intends to deploy a nuclear powered energy facility on the lunar surface by 2035 marks a decisive shift in human space activity. The International Lunar Research Station project, anchored by the so called Selena reactor programme, is not merely a technological escalation. It is the clearest signal yet that permanent, high risk infrastructure will soon operate beyond Earth under legal frameworks drafted for an era of slide rules and analogue telemetry.
Space law, as presently constituted, was designed to prevent territorial conquest and weapons of mass destruction in orbit. It was not designed to regulate autonomous decision making systems, cross border biometric data processing, cyber security threats to nuclear installations, or multinational commercial exploitation of extraterrestrial resources. The convergence of artificial intelligence, nuclear energy and permanent lunar presence has created a jurisdictional vacuum that no existing treaty regime can credibly fill.
At the centre of this legal crisis lies the Outer Space Treaty of 1967, still the constitutional charter of international space governance. Its core architecture rests on two principles that are increasingly irreconcilable in practice. Article II prohibits national appropriation of celestial bodies. Article VIII grants states jurisdiction and control over space objects and personnel registered under their authority. This duality functioned tolerably when activities were temporary and scientific. It collapses when infrastructure becomes permanent, autonomous and economically strategic.
A nuclear facility operated by Russian controlled artificial intelligence on the Moon would, under Article VIII, fall within the jurisdiction of the Russian Federation. Russian criminal law, civil law, administrative regulation and data protection statutes would attach to the installation and its personnel. At the same time, Article II declares that no part of the Moon may be subject to national sovereignty. The legal fiction that jurisdiction can exist without territorial sovereignty becomes increasingly untenable when installations operate continuously for decades, process vast quantities of data, and support industrial activity.
This contradiction produces what may be described as a legal enclave phenomenon. The territory remains formally international, but the operational reality becomes functionally national. It is a model already familiar in offshore drilling platforms and military bases, but with far greater consequences because no host state exists to contest jurisdiction.
The data protection implications alone are extraordinary. Russian Federal Law No. 152 FZ requires that personal data of Russian citizens be processed using databases physically located within Russian territory. If Russian engineers, scientists or contractors work at the lunar station, their biometric, health and performance data would fall under this statute. To comply, Russia would either need to replicate terrestrial data localisation infrastructure in space or classify the lunar facility itself as Russian territory for regulatory purposes.
Either approach undermines the legal character of the Moon as the province of all mankind. What begins as compliance with domestic privacy law becomes de facto sovereignty through data jurisdiction. Other states will inevitably respond by extending their own regulatory regimes extraterritorially to their nationals in space. The result is a patchwork of overlapping national legal systems operating on the same lunar surface, without a superior judicial authority to resolve conflicts.
The liability regime is equally fragile. The 1972 Convention on International Liability for Damage Caused by Space Objects establishes absolute liability for damage on Earth and fault based liability for damage occurring in outer space. Fault presupposes human agency. Negligence, recklessness and intent are legal constructs rooted in human cognition.
An autonomous reactor management system driven by machine learning does not fit this framework. If an artificial intelligence optimises cooling cycles, power distribution or radiation shielding in a manner that later causes contamination of neighbouring installations, lunar habitats or scientific equipment, how is fault established. The state may argue that no human acted negligently. The algorithm behaved unpredictably. The harm was emergent, not intentional.
Such arguments would exploit a structural weakness in the liability convention. Without reform, states could externalise catastrophic risk while internalising economic benefit. The incentive to deploy increasingly autonomous systems would rise precisely because legal accountability becomes diffuse.
Article VI of the Outer Space Treaty provides that states bear international responsibility for national activities in outer space, including those conducted by non governmental entities. Yet responsibility without an operational liability mechanism is largely symbolic. There is no international space court, no mandatory arbitration framework, and no enforcement apparatus capable of compelling compliance beyond diplomatic pressure.
The danger is not hypothetical. Nuclear installations are cyber physical systems. Their safety depends as much on software integrity as on mechanical engineering. A malicious intrusion into the control architecture of a lunar reactor could trigger physical consequences indistinguishable from a design failure.
Such an incident would constitute harmful contamination under Article IX of the Outer Space Treaty, which obliges states to avoid adverse changes to the environment of celestial bodies. Yet the treaty provides no technical standards, no inspection regime, and no enforcement authority. It merely requires consultation, a diplomatic procedure ill suited to crisis management when radiation dispersal is measured in seconds.
This is where space law collides with information security law. On Earth, nuclear facilities are subject to extensive cyber security regulation, including mandatory penetration testing, classified network architecture standards and continuous monitoring by national security agencies. None of these obligations exist in binding form for extraterrestrial installations.
The emerging reality is that the Moon is becoming a critical infrastructure environment, analogous to energy grids, undersea cables and satellite constellations. It is simultaneously physical and digital. A vulnerability in code is a vulnerability in matter.
Yet space treaties remain focused on rockets, orbits and flags.
The commercial dimension compounds the problem. If energy from the Selena reactor supports mineral extraction, fuel production or data centres servicing Earth based artificial intelligence systems, the Moon becomes integrated into terrestrial supply chains. Contract law, insurance law, competition law and securities regulation will follow.
Any multinational corporation partnering in such projects will be required to disclose material legal risks to shareholders. These include treaty instability, absence of predictable dispute resolution, sanction exposure, potential classification of activities as unlawful appropriation, and environmental liability without jurisdictional clarity.
Insurance markets are unlikely to provide coverage for nuclear operations in a jurisdiction where liability rules are undefined. Reinsurers already exclude war, cyber warfare and nuclear incidents in many contexts. Lunar nuclear risk may be entirely uninsurable, transferring exposure directly to corporate balance sheets and ultimately to states.
Geopolitically, the project intensifies strategic fragmentation. The Artemis Accords led by the United States propose safety zones around installations to prevent interference. Russia and China reject this model as disguised territorial appropriation. Without consensus, competing legal doctrines will coexist, enforced by technological capability rather than law.
This is how lawlessness emerges in advanced form. Not through absence of rules, but through incompatible rules applied without hierarchy.
What is required is not another aspirational declaration, but a new generation of binding instruments. A revised liability convention establishing strict liability for autonomous systems. A space data protection protocol harmonising jurisdiction over personal and operational data. Mandatory cyber security standards for off Earth nuclear facilities enforced through inspection mechanisms. An international space infrastructure authority with technical audit powers. And a permanent dispute resolution body with compulsory jurisdiction.
None of this exists.
Until it does, the deployment of artificial intelligence controlled nuclear infrastructure on the Moon will operate in a legal twilight zone. Engineers may celebrate innovation. Strategists may celebrate leverage. Lawyers, however, recognise the architecture of future conflict.
The Moon is no longer a distant scientific outpost. It is becoming a jurisdictional battlefield where code, atoms and sovereignty intersect.
If law does not evolve at the same pace as technology, then the first permanent human structures beyond Earth will not be governed by legal certainty, but by power, capability and unilateral interpretation.
History suggests that when law lags behind technology at moments of civilisational transition, the cost is measured not in court filings, but in decades of instability.
The lunar age is approaching. Whether it will be ruled by treaties or by technical dominance remains an open legal question.