COSMOS

I. The Evening Musk Decided

In the autumn of 2022, while the Ukrainian army was planning a decisive thrust towards Sevastopol, Elon Musk switched off Starlink. Not for everyone — only in the area where the Ukrainian drones were operating. Hundreds of drones, aimed at the starting positions of the Russian Black Sea Fleet, lost their connection. The attack failed. Musk later explained that he had wanted to prevent nuclear escalation.

Whether that decision was wise or foolish, morally defensible or strategically catastrophic — that is arguable. What is not arguable: Ukrainian military operations hung by a single switch held in the hands of a private entrepreneur. A man with no democratic mandate, no accountability to the Ukrainian people, who by his own account was simultaneously in contact with Russian Foreign Minister Lavrov.

This is the new reality of space: satellites decide the outcome of wars. And the most strategically critical satellites belong not to states but to corporations.

II. The First Threat: ASAT Weapons and the Kessler Syndrome

On 15 November 2021, Russia destroyed its own satellite Kosmos 1408 with a ballistic missile. The test generated more than 1,500 trackable debris pieces and hundreds of thousands of smaller fragments — and forced the crew of the International Space Station into an emergency evasive manoeuvre. The debris field is still in orbit today. It will remain there for decades.

In 2007, China had destroyed its weather satellite Fengyun-1C in a similar test, generating more than 40,000 debris pieces larger than one centimetre — the largest single contribution to space debris in the history of spaceflight to that date. In August 2024, a Chinese Long March 6A rocket fragmented in orbit, creating a new debris cloud that scientists described as a paradigm shift in risk assessment. In November 2025, China's Shenzhou-20 spacecraft was struck by a debris piece — the extent of damage was not disclosed.

As of 2025: approximately 9,500 active satellites and 36,000 tracked debris fragments in Earth orbit, plus an estimated hundreds of millions of smaller, untracked particles. Each of these moves at approximately 28,000 kilometres per hour. An aluminium fragment one centimetre in diameter strikes with the energy of a hand grenade.

The Kessler Syndrome — named after NASA scientist Donald Kessler, who described it theoretically in 1978 — is the scenario in which debris in low Earth orbit reaches a critical density and triggers self-perpetuating collision cascades. Each collision generates new debris that triggers new collisions that generate new debris. The result would be that low Earth orbit would no longer be usable for generations — perhaps centuries. Statistical models show that critical thresholds have already been exceeded or approached in certain orbital shells.

The decisive feature of the Kessler Syndrome is its impartiality. Debris does not ask what country a satellite belongs to. An ASAT test that destroys a hundred enemy satellites and generates a thousand debris pieces endangers the attacker's own satellites with those very same pieces — and the satellites of all other states. Those who launch a kinetic ASAT attack are building the weapon that destroys itself.

That is not a metaphor. That is physics.

III. The Second Threat: A Billionaire's Thumb on the Switch

In 2025 SpaceX operated approximately 6,000 active Starlink satellites in low Earth orbit — more than all other nations and companies combined. In Ukraine at that time there were an estimated 200,000 terminals. Hospitals, government agencies, drone operations, artillery coordination, intelligence transmission — everything ran over Starlink. Musk said it himself: without Starlink, the entire Ukrainian front would collapse.

This raises a question for which the twentieth century had no category: what is a private company that decides over the strategic infrastructure of a war? Not a state, not an ally, not an international organisation. A joint-stock company with a majority owner whose decisions are not controlled by democratic institutions.

Musk had in 2022 switched off Starlink in the operational area, thereby preventing a Ukrainian attack. He had at the same time, by his own account, been in contact with the Russian Foreign Ministry. And he had decided not to activate Starlink over Crimea — a militarily critical decision, taken without mandate, without debate, without accountability.

Europe has IRIS² in planning — a European satellite constellation in low Earth orbit, conceived similarly to Starlink. The first consortium of eight companies has a contract. Estimates for operation: at earliest 2030, realistically 2032. Until then, Europe — and Ukraine — is dependent on SpaceX. On a company whose owner simultaneously serves as an unofficial adviser to the US government, maintains close ties to the Chinese economy, and whose stated strategic priorities are oriented towards humanity's intergalactic project, not the tactical interests of a European ally in a ground war.

IV. The Third Threat: Invisible Dependencies — GPS and the Financial System

Satellites are not only indispensable for communication and navigation. They are the invisible foundation of global financial infrastructure. Banks worldwide synchronise their transaction timestamps via GPS signals. The SWIFT standard requires timestamp precision at the millisecond level — not achievable without global satellite synchronisation. A disruption of the GPS system would not only blind navigation systems and combat drones. It would destabilise global payment transactions.

GPS is a US military system. It can theoretically be switched off regionally — a capability the US military possesses, and for whose conditions of deployment no public legal framework exists. Europe has Galileo — its own global navigation satellite system, fully operational since 2016. But most European devices use GPS as their primary system and Galileo only as backup. The transition to Galileo as the primary system has never been decided politically.

The connection between COSMOS and AGORA is particularly close here: the financial sovereignty of Europe for which AGORA argues presupposes a sovereign satellite and timing infrastructure. You cannot operate an independent financial architecture when the time base on which it runs is controlled by a foreign military.

V. What the Outer Space Treaty Does — and Does Not

The Outer Space Treaty of 1967 is the foundation of space law. It prohibits the stationing of weapons of mass destruction in orbit, declares space the common heritage of humanity, and excludes national sovereignty claims over celestial bodies. For 1967, that was revolutionary.

But the treaty does not prohibit conventional anti-satellite weapons. It contains no regulation for kinetic ASAT systems, no debris liability, no sanctions for space tests that endanger the orbit for all. The UN General Assembly adopted a resolution in 2022 calling for a moratorium on direct kinetic ASAT tests — endorsed by 37 states, legally non-binding, not signed by Russia or China.

There is no equivalent to the Nuclear Non-Proliferation Treaty for space. There are no liability rules for debris attributable to a causing state. There is no international authority that manages access to orbit or sanctions debris-generating activities. The orbit is a common without governance.

That is the actual problem. Not the absence of satellites — but the absence of rules for a space on which the global economy, global navigation, the financial system and twenty-first century warfare equally depend.

VI. COSMOS — The Concept

COSMOS stands for Convention on Space as an Open and Maintained Strategic Commons. The name takes the core principle of the Outer Space Treaty seriously — space belongs to everyone — and adds the decisive second condition: it must also be kept usable for everyone.

The concept has three dimensions.

First dimension: The Kessler Principle as legal framework. The physics of the Kessler Syndrome provides the strongest self-punishment argument of the entire series: those who launch a kinetic ASAT attack generate debris that endangers their own satellites — inevitably, automatically, without any political decision. But this is not sufficient as deterrence, because short-term military calculation ignores the long-term self-harm. COSMOS translates the physics into law: every debris-generating activity in orbit — kinetic ASAT tests, deliberate collisions, negligent de-orbiting practices — is treated as an attack on a common good and triggers economic consequences proportional to the debris mass generated. Not as retaliation, but as compensation to all users of the orbit. Kessler liability makes destroying satellites so expensive that it no longer pays.

Second dimension: Governance of private space infrastructure. The Starlink problem is a legal vacuum: private infrastructure fulfilling the strategic functions of a sovereign state is subject to no corresponding obligations. COSMOS proposes a new category: Strategic Space Infrastructure — defined as any satellite system demonstrably integral to a state's military or critical civilian infrastructure. Such systems are subject to special obligations: no unilateral shutdown without 30 days' notice in peacetime. No politically motivated restrictions without transparency towards the affected state. No simultaneous communication with warring parties without informing the ally. The alternative is European sovereign capacity — which COSMOS sets as a bridging rule until IRIS² is operational.

Third dimension: European satellite sovereignty. Europe has the building blocks: Galileo (GNSS), Copernicus (Earth observation), IRIS² (LEO communication, under construction). What is missing is the political decision to treat these systems as strategic infrastructure — with corresponding investments, priorities and protection rules. COSMOS defines three concrete goals for the EU: Galileo becomes the mandatory base infrastructure for all European critical systems — banks, power grids, military — with GPS only as backup. IRIS² receives the status of a European defence project with corresponding prioritisation and financing. And the EU negotiates reciprocity clauses with its partners: strategic satellite services used from EU territory are subject to European law regarding availability and restrictions.

VII. The Kessler Principle — The Strongest Deterrence Mechanism of the Series

In the other concepts of the series, the self-punishment structure is political in nature: NUET excludes the nuclear weapons user from the world economy — a human decision. RIEGEL makes the Suwalki operation self-destructive through geographic counter-symmetry — a construction of geopolitics and topography. SHADOW makes transponder-dark navigation expensive through rules and databases — a bureaucratic instrument.

The Kessler Principle is different. It is not a political construction. It is physics. Whoever kinetically destroys a satellite generates debris that with the probability of a natural law hits further satellites — including their own. There is no evasive option, no room for negotiation, no political intervention that collects the debris back. Russia contaminated its own orbit in 2021. China contaminated the orbit in which it operates in 2007. The physics punishes the attacker without anyone having to do anything.

COSMOS adds the legal and economic punishment to the physical one. Not: we will retaliate. Rather: the debris remains in orbit. You have destroyed a part of the common home. That has a price. A Kessler liability convention — similar to the oil liability conventions of maritime law — would charge every state that generates debris the demonstrated costs of collision avoidance, satellite failures and orbit clean-up. Proportionally, automatically, without a vote.

That is civilisational deterrence in its purest form: nature has already built the architecture. COSMOS builds the legal roof over it.

VIII. What COSMOS Is Not

COSMOS is not a disarmament treaty. It does not require any state to give up its anti-satellite capabilities. It is realistic: states will develop and possess space weapons as long as they believe they have strategic utility. COSMOS does not make possession expensive — it makes deployment expensive.

COSMOS is also not a continuation of the Cold War in orbit. It is not directed against China or Russia as such. It is directed against activities that endanger the orbit as a common good — regardless of the perpetrator. The United States conducted ASAT tests in 1985 and 2008. India in 2019. All generated debris. COSMOS applies to all equally.

And COSMOS is not a European fantasy of technological autarky. Europe will remain dependent on foreign capacities for a long time. COSMOS defines rules for this dependency — no unilateral decisions on strategic services, transparency, reciprocity. Until the own capacity is in place.

IX. What Must Be Done

First: A Kessler liability convention, negotiated within the framework of a COSMOS agreement initially between EU and NATO states. Any demonstrated kinetic ASAT activity generating debris establishes international compensation obligations towards all affected satellite operators and states. The liability framework is modelled on the London Oil Liability Convention: automatic, proportional to the damage mass, without proof of intent. States that do not accede lose access to European space data and orbit coordination services.

Second: A category of Strategic Space Infrastructure in EU law. Satellite systems on which European states or the EU depend for critical functions are subject to availability obligations, transparency requirements and prohibition of restrictions without diplomatic advance notice. SpaceX can operate Starlink — but cannot unilaterally switch it off without notice over a European belligerent.

Third: Galileo as European mandatory base infrastructure. All European banks, power grids, mobile networks and military systems are required to use Galileo as the primary time synchronisation and navigation system. GPS remains backup. This is technically achievable — Galileo is operational and already integrated in many devices. Only the political commitment is missing.

Fourth: IRIS² receives the status of a European emergency project with corresponding financing and timeline. The current consortium of eight companies is not delivering fast enough. A partnership with Eutelsat/OneWeb as a bridging solution must be negotiated immediately to structurally reduce Starlink dependency before 2030.

Fifth: Active debris removal as a European programme. ESA's ClearSpace-1 is a pilot for active debris removal. COSMOS requires this to become a fleet programme — not because Europe has to clean up others' debris, but because the programme creates political credibility: Europe is the actor that maintains the orbit while others contaminate it. That is not weakness — that is a leadership position in the only space no nation owns.

Whoever destroys a satellite
destroys a piece of the orbit on which everyone depends —
including themselves.
Physics knew this since 1978.
Now the law must know it too.