Introduction

In 2023, researchers at Curtin University detected that 1,806 individual Starlink satellites, roughly 28% of the active constellation at the time, were leaking unintended radio emissions into protected radio-astronomy bands. Nobody violated a rule. Nobody could be held responsible. That is the problem this blog addresses.

When one constellation operating under a single national licence generates continuous harm across dozens of jurisdictions simultaneously, which state is responsible, under what legal standard, and to whom? The Outer Space Treaty’s (OST) framework was not built to answer that question. The International Court of Justice’s (ICJ) 2025 Advisory Opinion on climate change offers a due diligence template for beginning to fill that gap.

One Licence, Many Victims

Articles VI and VII of the Outer Space Treaty (OST) distribute international responsibility by reference to the “authorizing state”: the state that licences a private operator bears international responsibility for that operator’s “national activities” in outer space, and the “launching state” bears liability under the 1972 Liability Convention for damage caused by its space objects. The architecture is clean in theory, and it worked in the era that produced it. The Cosmos 954 incident, a Soviet satellite crashing onto Canadian territory, was a textbook bilateral harm: one state, one object, one victim.

Megaconstellations break every assumption in that model. The Starlink constellation now exceeds 7,000 satellites in orbit, with new launches occurring roughly every few days, all operating under a single Federal Communications Commission (FCC) authorization. The harm that constellation generates is not bilateral. It is continuous, orbit-wide, and felt by states that played no role in authorizing the activity. Botswana experiences degraded Global Navigation Satellite System (GNSS) accuracy. Australia loses Square Kilometre Array Low-frequency (SKA-Low) data. Neither state has a discrete event it can litigate under the Liability Convention’s fault-based framework for non-physical damage.

The Radio-Frequency Interference (RFI) Problem: A Gap in the Rules Themselves

The International Telecommunication Union (ITU) framework for radio-frequency interference is real, binding treaty law, but it regulates attributable interference, harm geographically locatable to a specific state’s territory. The Iran-Starlink dispute before the Radio Regulations Board (RRB) illustrates how that bilateral channel works and where it stops: Iran identified unauthorized Starlink terminals operating on its territory, attributed the interference to a geographically locatable source, and escalated through the ITU’s formal dispute mechanism. That process is workable only because the harm was bilateral and attributable. Unintended Electromagnetic Radiation (UEMR) from megaconstellations is a different problem: it originates in satellites’ electrical components, falls entirely outside existing ITU regulations and hits every radio-quiet zone on Earth simultaneously. The existing ITU framework does not clearly or adequately regulate UEMR, and where no clear rule applies, no breach can be attributed.

One might object that Article IX of OST already provides a textual foothold. The provision obliges states to conduct space activities “so as to avoid their harmful contamination” of outer space, and the Cologne Commentary on Space Law (Volume 1) (p. 573) indicates that this contamination clause should be read from the perspective of interference with outer space research activities, not merely physical environmental harm. On this reading, UEMR leaking into protected radio-astronomy bands could plausibly constitute “harmful contamination” within Article IX’s scope. The difficulty is that Article IX’s contamination obligation is framed as a duty of conduct in planning and conducting activities, not as a strict prohibition with a defined threshold. Without agreed technical standards specifying when electromagnetic leakage crosses the line into “harmful contamination,” the provision remains a general duty of vigilance rather than an operative primary rule from which breach can flow. The provision functions as a framework norm that requires further specification before it can generate concrete obligations of the kind necessary for state responsibility to attach. This is precisely the gap that due diligence, as an objective standard of conduct tied to available scientific knowledge, is equipped to fill.

Why Article VI Cannot Absorb This

Even where an operator’s conduct might be characterized as failing the “due regard” standard under Article IX of the OST, Article VI ties responsibility to the authorizing state alone. That state, typically the United States for Starlink, bears responsibility to other states parties generally, not to any specific injured party. The result is that an injured state has no direct claim against the licensing state under the OST itself, and the Liability Convention’s fault-based regime for non-physical harm is, as the literature on Article VI responsibility confirms, essentially undeveloped.

The ‘due regard’ obligation in Article IX may, however, carry more legal weight than this suggests. The Cologne Commentary draws an explicit parallel between Article IX’s ‘due regard’ clause and its counterpart in the United Nations Convention on the Law of the Sea (UNCLOS) Article 87, where the obligation has been understood as encompassing a state’s responsibility for the acts and omissions of non-state actors operating under its jurisdiction that hamper the rights of other states to use the high seas. On this reading, a state that authorises and supervises a mega-constellation operator, but fails to take reasonable steps to prevent that operator from interfering with other states’ rights to conduct space activities may already engage responsibility under Article IX itself, not merely under Article VI’s ‘national activities’ model.

The deeper structural problem is that the ‘national activities’ model assumes injury is traceable to a single bilateral relationship. Mega-constellation harm is instead structurally erga omnes-like: it affects the international community as a whole. That framing points to the legal category the ICJ’s 2025 Advisory Opinion addresses.

What the Climate Advisory Opinion (AO) Actually Offers

The move here is not to suggest the AO solves space law. It is to test whether three specific structural elements are worth borrowing.

The erga omnes framing of diffuse harm

The ICJ held that obligations to prevent significant harm to the climate system are owed to the international community as a whole, not to any one bilaterally injured state. The Court at paras. 439 – 440 characterised these obligations as erga omnes under customary international law, grounded in the common interest of all states in the protection of global environmental commons, a framing not dependent on treaty membership. That framing applies directly to mega-constellation harm: there is no “the” victim state for orbital RFI, so the bilateral liability logic the OST silently assumes is the wrong tool regardless of which body of law we are in.

That erga omnes framing has direct consequences for state responsibility. Under Article 42(b)(i) of the Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA), a state qualifies as an ‘injured state’ where a collectively owed obligation is breached and that state is specially affected. As Gaja analyses in Crawford’s Law of International Responsibility, this provision addresses situations where the obligation is collective but one state bears a disproportionate share of the harm. If a licensing state’s failure to regulate its constellation breaches the collective obligation to maintain a non-interfering outer space environment under Article IX, then a specially affected state has locus standi as an injured state under Article 42(b)(i), not merely as a third party under Article 48, and may claim full reparation accordingly.

Due diligence as a standard of conduct, not a guarantee of result 

The ILC’s ongoing work on due diligence in international law, advanced through the First Report of Special Rapporteur Dr. Penelope Ridings, introduced before the Commission in June 2026, will generate a cross-regime framework that, once concluded, could supply the general-law scaffolding that an Article IX-based space law standard currently lacks.

Applied to the licensing state’s obligations under Article IX, a due diligence standard would operate as follows. The authorising state, the United States in the case of Starlink would not be required to guarantee that its licensee causes no UEMR or collision risk. It would instead be required to demonstrate that it took all reasonable steps, commensurate with the known and foreseeable risk, to prevent such harm. As the Ridings Report confirms across regimes, due diligence is a standard of conduct, not of result: what is required is sustained vigilance, proportionate to the magnitude of the risk. In concrete terms, this means the licensing state must impose monitoring and reporting obligations on its licensees, audit compliance with those obligations, and enforce technical mitigation standards as the science evolves. 

Rejection of the “too diffuse to attribute” defence 

Perhaps the most transplantable element of the AO is its treatment of cumulative, multi-state harm at paras. 429-438. The Court rejected the argument that collective harm is too complex to attribute to specific states, holding instead that scientific methods, data, and evidence can establish the causal link. Space Situational Awareness data plays the same evidentiary role for orbital harm that emissions-attribution science now plays for climate harm. “Orbiting Chaos” itself flagged Space Situational Awareness (SSA) data as a potential bridge between scientific detection and legal responsibility; the climate AO provides the doctrinal framework for treating that bridge as legally operative.

The limiting principle deserves emphasis: due diligence is an obligation of conduct. Not every leaked signal and not every close approach becomes an internationally wrongful act under this framework. Only a failure by the licensing state to regulate, audit, or enforce commensurate with known risk would engage responsibility. That is a high bar, but it is at least a bar, which is more than the current framework provides.

Where the Analogy Breaks Down

Intellectual honesty requires naming the limits, and there are three real ones.

The climate AO rests on decades of accumulated customary practice rooted in Trail Smelter, the ILC Draft Articles, and the Nuclear Weapons Advisory Opinion. Space law has no equivalent lineage for orbital congestion or RFI specifically. Article IX’s “due regard” language, as “Orbiting Chaos” itself noted citing Jarose, operates as a general conduct obligation requiring contextual balancing of interests. That is a thinner textual foundation than the UNFCCC and Paris Agreement gave the ICJ to work with. A due diligence obligation for mega-constellations would have to be grounded in Article IX alone.

The severity question also matters doctrinally. Climate harm is irreversible and existential in the way that drove para. 138’s “stringent” standard. The ICJ’s language at para. 138 is explicitly keyed to the “quintessentially universal risk” character of climate change. Orbital RFI, while scientifically costly, is not of the same order, and a lower standard of care may therefore be appropriate. 

Finally, the institutional machinery differs. The UNFCCC and Paris regime gave the Court treaty text to anchor Nationally Determined Contribution (NDC)-related due diligence obligations. The ITU Radio Regulations are silent on UEMR by design. Any due diligence obligation in the space context would have to be built interpretively, without the treaty scaffolding the climate regime provides. Para. 457 of the Climate AO, which addresses the relationship between general customary obligations and specific treaty regimes, is at least a potential bridge here, but it is a narrow one.

The stretch is still worth making. The alternative is leaving UEMR and multi-jurisdictional collision risk entirely outside binding international law, which is precisely the status quo “Orbiting Chaos” showed to be untenable for debris.

The Way Forward

Article VI and VII’s single-authorizer model was not built for activities whose harm is structurally multilateral. Due diligence, transplanted carefully from the climate AO, gives international law a vocabulary for holding authorizing states to a standard of conduct without requiring an identifiable bilateral victim and without demanding a treaty overhaul. The concrete next step is modest: a COPUOS-endorsed framework, or a resolution at the ITU’s World Radiocommunication Conference, explicitly adopting an SSA-evidenced due diligence standard for licensing states, obligating them to require their licensees to monitor, report, and mitigate demonstrable harm commensurate with what the science shows. The question that remains is whether states will choose to operationalise these tools within existing frameworks, or whether orbital governance will continue to rely on the inadequacy of silence.

---

Suhani Agrawal is a second-year student at National Law Institute University, Bhopal.

---

Picture Credit: James McTaggart / Midjourney