Neither. There is no right of way in orbit.
There is no traffic law in space. No mandatory avoidance duty. No binding standard of care telling an operator when it must maneuver, who must yield, or how close is too close.
Which means when two satellites collide, the legal question is not who violated the rule. It is whether any rule existed to violate.
This has already happened
In February 2009, the operational communications satellite Iridium 33 collided with the derelict Russian military satellite Cosmos 2251 roughly 490 miles above Siberia. Both were destroyed instantly. The collision generated thousands of pieces of trackable debris, much of which is still in orbit and still being avoided by other spacecraft today.
Nobody paid anybody. No claim was pursued to resolution.
That is the leading real-world example of a satellite collision, and its legal legacy is essentially nothing. Seventeen years later, the framework has barely moved.
The fault standard, and why it is nearly unusable
The Liability Convention of 1972 draws a line that decides everything:
| Where the damage occurs | Standard | Article |
|---|---|---|
| On Earth’s surface, or to aircraft in flight | Absolute liability. No fault required. | Art. II |
| In space, to another space object | Fault-based. Fault must be proven. | Art. III |
Damage on the ground is the easy case, relatively speaking. See who gets sued when space junk hits your house.
Damage in orbit requires proving fault. And here is the problem that has never been solved: fault requires a standard of care, and there is no binding standard of care for orbital operations.
To prove an operator was at fault, you would need to establish what a reasonable operator should have done. Reasonable by what measure? There is no regulation requiring a maneuver at a given collision probability. No mandated response time to a conjunction warning. No rule assigning responsibility between two maneuverable satellites. No requirement to maintain propellant reserves for avoidance.
Operators follow internal thresholds and industry practice. Those are voluntary, they vary widely between companies, and they have never been endorsed by any body with authority to make them binding.
An expert could testify to prevailing practice. Whether a tribunal would accept prevailing voluntary practice as the legal standard has never been tested, because no case has gotten that far.
What actually exists
Not nothing. Just nothing binding.
Conjunction warnings. The U.S. Space Force tracks objects and issues conjunction data messages to operators worldwide, predicting close approaches. The Office of Space Commerce has been developing a civil space traffic coordination capability to take over and expand this function.
These are warnings, not instructions. Nothing obligates an operator to act on one. An operator that received a warning and did nothing has a documented notice problem, which is genuinely useful evidence, but notice alone is not a breached duty if no duty existed.
Voluntary debris mitigation guidelines exist through international coordination bodies and national licensing conditions. Licensing conditions are the sharpest tool here: a U.S. operator’s license can carry requirements, and violating a license condition is much closer to actionable fault than departing from a voluntary guideline.
Insurance and cross-waivers. Much of this gets handled contractually. Launch participants routinely waive claims against each other, and satellite operators carry in-orbit coverage. That is why disputes rarely surface as litigation. It is resolved by contract and insurance, quietly, among sophisticated parties.
Who bears the loss, in practice
Follow the money rather than the doctrine.
The satellite owner’s insurer. In-orbit coverage pays, and this is the ordinary outcome. Whether the insurer then pursues subrogation against the other operator depends on how provable fault is, which brings you back to the standard-of-care problem.
The operator’s own balance sheet, for uninsured or self-insured constellations. With very large constellations, some operators effectively self-insure and treat losses as operational cost.
Downstream customers. A broadcaster, a telecom carrier, a government agency, or an imaging customer that loses service. Their claims are contractual against the operator, governed by service level agreements that usually limit liability significantly.
Everyone else, forever. This is the part nobody pays for. A collision creates a debris field that raises collision risk for every other operator for decades, and imposes real ongoing cost in propellant, avoidance maneuvers, and lost operational life. Nobody compensates that. There is no mechanism to.
Where a real claim could come from
Setting aside the treaty channel, which requires a State to bring the claim, there are two paths that could produce actual litigation in a domestic court.
Product liability. If a satellite failed to maneuver because a thruster, an attitude control system, or flight software failed, that is a defect claim against a manufacturer. Conventional product liability litigation, governed by ordinary law, with a familiar structure. Contractual limitations of liability in aerospace supply agreements are aggressive, but this is by far the most viable route to a real courtroom.
Contract and negligence between commercial parties. Two private operators, both U.S. entities, in a domestic court, arguing about a collision. No treaty required. The standard-of-care problem persists, but a court could resolve it, and the first court that does will effectively write the rule.
That is why this area is worth watching. The first serious commercial satellite collision litigation will create the standard of care that does not currently exist.
The questions with no answers yet
Does ignoring a conjunction warning establish fault? The most consequential open question. Untested.
Is a maneuverable satellite obligated to yield to a derelict one? Intuitively yes, since the derelict cannot move. No rule says so.
Who is at fault when both operators maneuver and make it worse? Two independent decisions, no coordination requirement, no communication protocol.
Does an autonomous collision avoidance system shift responsibility to its developer? Large constellations use automated avoidance. If the algorithm chooses wrong, that is a software defect question, and the same unresolved “is software a product” problem appears here as it does in autonomous vehicles.
Who pays for the debris field? Nobody. There is no mechanism, and as orbital density increases this becomes the defining unpriced externality of the industry.
What this means for you
If you operate a satellite, own a payload, or depend commercially on one, the practical protections available today are contractual and evidentiary rather than legal.
Document your conjunction response decisions and the thresholds you applied. If a collision occurs, your own records establishing that you followed a defined and reasonable process are the best evidence you will have, and the absence of that record is the easiest thing for an opponent to attack.
Read the liability limitations in your launch, supply, and service agreements as the operative allocation of risk, because they will govern long before any treaty does.
Robert Law Group is building a practice in emerging technology torts, and space is where the gap between activity and law is widest. Aimée Robert studied at the United States Air Force Academy and spent two decades in complex multi-defendant litigation against corporate parties. When the defense is “there was no rule requiring us to act,” that is an argument about the standard of care, and establishing a standard of care where none is written is exactly what complex product litigation does.
Call 832-509-2303.
Future Lawsuits. The law before the law catches up.