This is a follow-up to my December essay, "Towards a Global Earth Observation Accord." That piece sketched the case for political, not only technical coordination of civilian Earth observation (EO). Several events since December have convinced me that the Artemis Accords comparison is more useful than less, than when I first thought.
My last essay looked at the current EO governance frameworks that exist (CEOS, GEO etc.) at length. I will not repeat that diagnosis here. This is a forward-looking essay, building on what I had written before.
As of June 2026, the Artemis Accords – a US-led framework setting out principles for cooperation in lunar exploration – have 67 signatories.
Read that number again: sixty-seven countries. Many of them with no satellites in orbit and no specific lunar exploration strategy have signed a principle-based framework for cooperation in lunar activities, resource utilization, and orbital deconfliction. The Accords launched with eight founding signatories in October 2020 and have grown more than eight-fold since.
Now, let's look at EO: according to our research, at least sixty (60) countries are launching or have announced plans to launch national EO satellite programs in the last 5 years. Many of them have arguably far more direct economic and operational stakes in EO than they do in lunar resource extraction, at least in the short-term. And yet, for all that investment, there is no global framework that commits governments to sustaining or sharing the EO data they produce.
The Artemis Accords have produced a working framework for an activity most signatories are not doing. EO has more participants, more economic value, more public benefit, and a fraction of the political infrastructure.
What Happened in EO in 2026
The absence of a political framework is no longer abstract as four recent events have made the costs of that quite concrete.
Planet Labs' indefinite withhold. On April 5, 2026, Planet announced that, at the request of the US government, it would indefinitely withhold imagery across Iran, Iraq, Israel, Lebanon, and the Gulf states, retroactive to March 9. In other words, the company moved from open commercial access to a "managed access" model with case-by-case review. Other US providers, including Vantor and BlackSky, implemented similar restrictions. Customers diversified to using open data such as Copernicus and Landsat, when possible, and started switching to international commercial EO providers, based outside US jurisdiction.
The Washington Post's open-data investigation. While commercial access was tightening, the Post, working with researchers at Oregon State's Conflict Ecology lab, published a detailed analysis of structural damage across Iran using Copernicus Sentinel data. The investigation mapped strikes on US military assets in the Middle East using Sentinel data cross-referenced with Iranian state media. In essence, when commercial open access closed, public open data, in this case, from Europe's Copernicus, became the backbone of independent, jurisdiction-neutral reporting.
Switzerland opts out of Copernicus. On June 5, the Swiss Federal Council decided Switzerland would not participate in Copernicus for the 2028–2034 cycle, continuing its current opt-out decision. Despite the fact that the commissioned economic assessment had recommended joining on societal, strategic, and economic grounds, the decision went the other way on budget grounds. Under Copernicus's open-data policy, Switzerland retains access to nearly all raw data. What it loses is operational services, governance, and procurement. The structural question this raises sits with everyone else: open data is a political choice, not a natural law, and the more it can be enjoyed without contributing, the more pressure builds on the current contributor-heavy model that funds the data.
Global ocean observation overhaul. In the same week, two governments produced opposite signals about ocean observation infrastructure. The US National Science Foundation announced plans to dismantle instruments part of the Ocean Observatories Initiative, a deep-ocean monitoring network across the Pacific and Atlantic, following a funding cut. Not long after, the European Commission formally adopted OceanEye, a new ocean observation framework and called for an international alliance to coordinate global ocean observation, explicitly framing this as "essential infrastructure, on a par with meteorology and space-based observations."
Open EO Data Is A Political Choice
The case studies, taken together, point at a deeper structural shift. Landsat Next, the planned US continuity mission, is targeted for cancellation in the latest budget request and according to the NASA administrator, could be restructured to a commercial transition. The question Switzerland's decision puts plainly: open data is a political choice, and the contributor base that funds it is fragile too.
With Landsat Next descoped and probably moving to a commercial procurement-based model (details of which are unknown), the global civilian EO commons is now substantially underwritten by a single program – Copernicus, which is funded by a coalition of European countries, including non-EU members like Norway and the UK. In other words, the world's open EO infrastructure depends on those governments that choose to keep paying for it.
That arrangement worked when three conditions held: the cost curve was modest, contributor budgets were healthy, and the political consensus around civilian EO was firm. None of those three is true today.
Costs will rise as next-generation Sentinel missions start to come online – the Copernicus expansion missions, Sentinel-1 NG and Sentinel-2 NG . Not only do they cost to build and launch, but the demands on processing, storage, and dissemination also scale with them. Fiscal conditions are tightening across Europe simultaneously. The question that Switzerland's decision raises: open access is a political choice, not a natural law.
How long this concentration holds is the structural question the next decade has to answer. Europe's commitment to EO data continuity is real and welcome, and it looks like they are keen to continue their investment as we see next-generation satellites get funded. But continuity carried by one regional coalition is not the same as durable global public-good infrastructure.
I am not trying to make an idealistic case for global cooperation – I know this is 2026 and the age of sovereignty. I am trying to make a case for reasonable risk-management. Every commercial EO company could eventually face Planet's situation. Every public program is one political shock from a Brexit-style problem — Copernicus itself nearly lost CO2M, the greenhouse gas monitoring mission, when the UK's exit created a €750M funding gap in 2021 (the UK rejoined Copernicus later in 2023).
When many nations each commit their baselines to a shared pool, the commons stops depending on any single program. Users get more open data, and no single decision can break it easily. This is resilience by spreading the load, not by making any one source replaceable. Losing a data source would still hurt, but a distributed base survives shocks a single-funder one cannot.
EO needs a better political framework and global governance and we need to have conversations (sometimes hard ones) to see how we can achieve something like that.
What the Artemis Accords Got Right
Before going further, it is worth being precise about what the Artemis Accords actually achieved, and what they did not.
The Accords are not a treaty, rather a series of bilateral memoranda of understanding, anchored in the existing Outer Space Treaty, that each signatory signs individually with the United States. They are non-binding, and they do not establish enforcement, dispute resolution, or liability mechanisms. By any conventional standard of international law, they are weak instruments. And yet, we have 67 signatories.
The features that produced that result are not lunar-specific. They could very well be templates for getting countries to commit to norms in an emerging domain.
Principle-based, not treaty-based. The Accords do not require parliamentary ratification or transfer of sovereignty. You read the principles, you decide they reflect your values, you sign. The political cost of joining is dramatically lower than for any UN-led framework, and that lower cost is the entire reason 67 signatures exist.
Bilateral with shared text. Each country signs the same document with the United States, individually. This avoids the lowest-common-denominator drag of multilateral negotiation and produces a recurring news cycle. Basically, it is a signing ceremony with each new partner, photographs and press releases included.
Diplomatic value beyond participation. Most signatories are not building lunar landers and have no plans to so far. They sign because the Accords offer alignment with a US-led coalition, signaling about which side of an emerging norms divide they are on, and a credentialed entry into space diplomacy. The signing is the product.
Incremental commitment. Signing does not require spending money or building hardware. The commitment is only to the principles. The hard implementation work such as interoperability standards, deconfliction and data sharing happens in workshops and bilateral conversations after. Artemis got the order right: political momentum first, technical detail second.
A specific convener with a brand. NASA and the US Department of State, in a defined ceremonial format, act as a magnet. There is a clear front door. A new country knows who to call. The Accords have a website, recurring meetings and a steady cadence of additions that compounds into permanent diplomatic infrastructure.
Most of these features have nothing to do with the Moon. They are a template for political norm-setting in emerging domains. And, that template is exactly what EO governance lacks today.
Meteorological Data: The Working Precedent
There is, however, one place where political coordination for satellite data already exists and works pretty well globally across all axes of power: weather.
The World Meteorological Organization (WMO) has coordinated global meteorological data sharing for decades. Its foundational data-sharing resolutions established the principle that essential meteorological data flows freely between members for weather forecasting purposes. A real, operational political layer exists, and it has moved petabytes across borders for years.
So, it might be also tempting to romanticize. Weather data sharing was not always open and was not always shared. The US Defense Meteorological Satellite Program was highly classified for most of its operational life. India's INSAT had selective access restrictions. European meteorological and mapping authorities sat under defense ministries for much of the twentieth century. The current openness of meteorological data is the result of a long political project — roughly fifty years of deliberate institutional construction, often against initial resistance from the defense establishments that ran the original satellite programs. Anyone claiming weather data has "always" been shared is forgetting a part of history.
That historical correction matters because it sharpens both the existence proof and the challenge.
- The existence proof: A political layer for satellite data coordination has been built before. The specific artifacts such as political commitments on data flow, reciprocal exchange architecture, coordinated continuity, frameworks anchoring commitments to specific observations do exist and operate today. They are not theoretical. The question for EO is whether the same artifacts can be built, not whether they could exist at all.
- The challenge: EO faces sharper constraints than weather did. The dual-use shadow runs through the entire industry in a way it does not for weather – a humidity sounder has limited tactical value, but a sub-meter optical image is a different matter. Economic universality is weaker as every country has weather, but high-resolution imagery over your own territory has different value than imagery over someone else's. The political starting position is harder. EO is roughly where weather was in the 1970s in terms of governance maturity.
Weather coordination took decades, but we may not have fifty years here.
What this means structurally: the weather toolkit transfers, but the weather timeline does not. EO needs the institutional artifacts from weather, on something closer to the Artemis timeline. It was built once before, so they can be built again, faster, scoped to EO's harder constraints.
The Dual-Use Problem in EO
This is the place to be honest about why an Accord like this does not already exist for EO.
First, an 'EO Accord' for civilian use assumes there's a clean line between civilian and defense data – but there isn't. We have the same sensors, same operators, and same data products for both. Basically, different uses applied to the same data.
The Planet case is the cleanest demonstration: the imagery that humanitarian organizations used for disaster response was the same imagery that intelligence analysts used for battle damage assessment, was the same imagery that insurance underwriters used for parametric flood payouts, was the same imagery the US government eventually wanted withheld because it was strategically inconvenient. There is no clean line through that data flow.
The EU's own 2026 policy architecture reflects this. The emerging European EO stack explicitly separates Copernicus (the civilian commons) from the proposed EOGS (Earth Observation Governmental Service, dual-use) and from ESA's European Resilience from Space programme. The institutions that built civilian EO are formally acknowledging that the category needs internal differentiation.
The "civilian use" framing is a label applied to a use case, not a property of the data.
Dual-use nature of EO is also the structural reality underneath every existing limits of EO governance. The existing bodies such as The Group on Earth Observations (GEO) and Committee on Earth Observation Satellites (CEOS) work around this problem rather than through it, by limiting their mandates to technical coordination that does not require taking a position on what counts as civilian. The minute you ask for stronger political commitments, dual-use becomes load-bearing.
Any honest proposal has to address this directly. An EO Accord that claims to govern "civilian EO" as a clean category will not survive contact with operators who derive fifty to eighty percent of their revenue from defense customers, with constellations explicitly designed for dual-use, or with the simple fact that the same Sentinel images that the Post investigation used for accountability journalism was the kind of data the US government would prefer to restrict in some contexts and rely on in others.
The Limits of Coordination
Now that we have addressed the dual-use scoping, what does an Accord actually do that the existing setup does not?
First, as I argued in my last essay, the current institutional layer is quite technical in nature: CEOS coordinates missions and standards, GEO promotes open data sharing and interoperability and the UN Disaster Charter activates emergency data flows on request. These bodies do necessary work, and the EO community is better for their existence. But they share a fundamental design constraint: they are technical and consultative by mandate. They were built to coordinate, not to commit.
This is not a critique. It is a structural observation about what they were built to do and what they were not. When a civilian mission gets cut due to budgetary challenges leading to important data gaps globally, there is no process for the pieces to be picked up by another country, unless via bilateral agreements. When Sentinel-1B failed in 2021 and global SAR coverage degraded for a few years, we had no framework to identify other datasets that could fill the gap (we all just took the hit).
A proposed EO Accord would operate at the layer the existing bodies were not designed to fill: making norms explicit and committing governments at the ministerial level rather than working groups. This is the layer the existing bodies do not occupy. It is not in competition with them, rather it sits above them. CEOS and GEO could continue their ongoing activities, ideally accelerated by the political commitments an Accord would create.
What It Can Actually Govern
It cannot govern "civilian EO" broadly, because that category is unstable. What it can do is narrower: signatories list specific datasets they commit to keep open, maintain continuity for, and protect from arbitrary restriction.
Most of the Copernicus and Landsat data would be part of it. Archives of EO data from Indian, Japanese, and Korean open archives would be in it. Data from a number of satellites launched by the 60 countries could be part of it. Some specific datasets from commercial EO companies could be listed if they have an open data policy. Obviously, most defense-focused data would not be in it.
But, it cannot resolve the dual-use ambiguity at the operator level. Planet will continue to serve both humanitarian customers and defense intelligence. ICEYE will continue to grow on defense revenue while providing SAR data for disaster response. The Accord does not change the underlying business models, and it should not pretend to.
It cannot prevent designated civilian data from being used for defense purposes. The Post investigation used Sentinel data — civilian, open, EU public infrastructure — for mapping war damage. It is open data behaving as designed: available to journalists, researchers, humanitarians, and to anyone with internet access including parties to the conflict. The Accord cannot and should not try to constrain who reads open data.
What it can do is create predictable rules for that designated baseline. That is a narrower claim than "global civilian EO governance," and it is the only claim that survives the dual-use problem honestly. Everything that follows in this essay should be read in that light: designated baseline, political commitments to maintain it and predictability around how restrictions affect it.
Don't think of what I am proposing as a governance mechanism for a clean civilian–defense line that exists in EO (it does not). Think of it as a defined list of EO datasets that signatories commit to protecting, because each of them wants a baseline no single government or company can switch off.
The realistic objective of an EO Accord is narrower: a principle-based political framework through which governments designate specific EO datasets as part of a global public baseline and commit to transparency, continuity, interoperability, crisis access, and provenance around them.
What an EO Accord Could Mean
In practice, the political layer is a list of commitments. Six are worth proposing, ordered here by accessibility — the first four within reach of any signatory, the last two scaling with operational capacity.
- Observation transparency
Each signatory commits to publishing what observations its EO missions deliver — sensor types, spatial and spectral resolutions, revisit rates, geographic coverage, and data access terms — and to keeping that picture current. This is the most accessible commitment and costs almost nothing. And yet, anyone trying to assemble a global picture of available observations for a specific use — say, 3m multispectral coverage for crop monitoring — has to cross-reference national space agency sites, CEOS records, and operator pages by hand (or hire TerraWatch to do it for them). The global EO community has no standardized, accessible picture of what is being observed by whom.
- Civilian–dual-use designation
Each signatory declares, at registration, which of its EO assets are designated as civilian baseline — and thus subject to the rest of these commitments — and which are out of scope. This is the operational form of the dual-use honesty above. Signatories are not asked to open everything, rather just asked to be explicit about what could be open.
- Interoperability standards
New missions commit to alignment with established formats, calibration networks, and metadata standards as a condition of contributing to the global baseline. CEOS and other technical bodies can define the standards while the Accord politicizes the commitment to follow them. This matters most for the new programs, where every non-standard launch is interoperability debt that someone has to pay later.
- Crisis data sharing
In declared emergencies such as natural disasters or humanitarian crises, designated civilian baseline EO data from their programs is made available to a global data-sharing pool by default to affected populations and humanitarian responders, without case-by-case approval. This is broader than the current Disaster Charter, which is activation-based and limited. It is a standing commitment, not an exception.
- Provenance and integrity
Designated baseline EO data carries verifiable provenance, with technical standards for cryptographic signing, metadata, and unalterable archiving. This is becoming non-negotiable in an era of synthetic imagery. Cheap to commit to and increasingly important to the credibility of the entire data flow.
- Continuity of foundational records
Signatories operating or have plans to operate long-term missions such as Landsat or Copernicus-equivalent commitments can declare they commit to either sustaining them, providing advance notice of termination, or coordinating handover to successor capacity. This is the heaviest commitment, but most likely only applies meaningfully to a smaller set of signatories. But it is the one that prevents the next Sentinel-1B-style gap.
Why This Could Scale
These six commitments don't weigh equally on signatories, and that is the design.
The asymmetric structure matters. Almost any signatory can meaningfully commit to transparency, designation, interoperability, and provenance — the first four. Crisis sharing depends on operational capacity, while continuity falls heavily on a smaller set of countries with plans for long-term missions. This, I think, is the design feature that makes the Accord broad rather than oligarchic. A country with a single observation satellite can sign and immediately be agreeing with the core of the Accord. This is how Artemis got smaller countries to sign alongside the major space powers: a low threshold for entry, asymmetric substantive contribution.
And it is the answer to the question the sixty-plus countries raise. Right now, more national EO investment is producing more fragmentation, not more capability. Every new program launched without shared norms is a small subtraction from the global baseline, not an addition.
Sixty countries launching more EO satellites could mean 60 contributions to a global Earth observation system or just 60 isolated national EO programs that do not add up.
The Accord is the political mechanism that makes them add up. Without it, the 2030s are a more fragmented EO landscape than the 2020s, not a richer one.
This is also where "Copernicus or nothing" stops being acceptable. Right now, when something matters globally – disaster response, war monitoring, climate monitoring — there is one program that reliably delivers at scale and openness and one that is guaranteed to continue fully into the 2030s (I cannot say the same about Landsat with certainty, which seems to be going through a transition). I think that is a precarious position for the world to be in, regardless of how good Copernicus is. The sixty-plus new programs could be the answer, but they will not be unless someone builds the framework that makes them add up.
The Private Sector Question
The Accord proposed above commits governments to a global public EO infrastructure. But let's not forget: the EO landscape today is dominated by the private sector. Not acknowledging them would be incorrect as we look to the future of EO.
But, in reality, only two questions matter: what role does the private sector have in this Accord, and what the Accord does when commercial restrictions happen to users anyway.
Why EO Companies Cannot Lead This
The entire commercial EO industry is increasingly defense-revenue-dependent. Planet reported approximately 18% commercial revenue in 2025 — meaning roughly 80% came from government, majority of which from the defense and intelligence segment. Other EO companies derive roughly three-fourths of their overall revenues from defense and dual-use contracts. ICEYE's, the new decacorn, growth story is overwhelmingly defense. This is the structural reality I have written about elsewhere (The Defense Paradox in Earth Observation) and it is not going to reverse.
What this means for the Accord: signatories must be governments, not companies. EO companies making most of their revenue from defense customers cannot meaningfully commit to something that requires them to be open and conflict with their main customer's behaviour. The Accord operates at the government commitment layer, while commercial operators are a regulated layer beneath it, not part of it.
Where The Private Sector Fits
That does not leave the private sector outside the Accord, rather it makes them the implementation partners. Governments are no longer building every satellite they need, rather they buy the capability from EO companies. The baseline the Accord commits to will be assembled as much from acquired commercial data as from public missions. The real question is what reason they have to feed the pool rather than hold their data back (as they do currently).
The Accord changes the terms of the buy, not the fact of it. A normal government purchase is restricted-use and sits behind a contract. A commitment to maintain an open baseline turns that into two things a supplier actually wants: a predictable, repeatable buyer instead of one-off procurement, and reach. Data sitting in one discoverable pool gets found, sampled and trialled by far more users than any company's own archive does, and that trial-to-paid path is the top of the EO sales funnel. The few companies already running open-data programs – Planet, Umbra, Wyvern – are doing a smaller version of this on their own, because the exposure is worth it. The Accord generalises that incentive to the whole sector.
It is also how continuity gets filled: when a public observation lapses, the cheapest fix is usually to acquire the commercial equivalent into the same committed pool rather than fund a new public mission. And none of it asks a company to sign anything or to be open by default. It asks them to sell into a baseline that pays them back in demand and reach. That is the role: not signatory, not convener, but the implementers of what the Accord asks.
The Global EO Accord As a Back-Up
The Accord does not and should not bind commercial operators, nor can it prevent Planet-style restrictions by governments. What it can do is make those restrictions less consequential for users.
When Planet shut down imagery access across the Middle East in April, customers were left to figure out alternatives on their own. What other 1-meter data was available? Were there EO companies from Europe, Asia or Middle East as alternatives? Today, those questions take manual research and produce uncertain answers (I am hired to do some of these jobs and I can tell you that the conclusions are are not very straightforward).
A global EO Accord could potentially change that. By providing a known list of committed datasets — with stated commitments around continuity, openness, and crisis access — it gives users an immediate backstop to consult when commercial access fails. The transition from a restricted source to alternative committed sources becomes navigable rather than chaotic.
A committed list tells users which alternatives exist and on what terms, but it does not promise any of them replaces what was lost. If the shut-off source was sub-metre tasking (like from Planet), another data source may still not match its resolution, revisit, or ability to task on demand. The backstop removes the scramble, not the shortfall. It is a coordination layer beneath the commercial one, not a capability guarantee.
So, the point of the is not to fix the commercial conduct problem – a Planet-like situation may happen again in the future. We cannot try to constrain shutter control authority. We cannot directly address closed-by-default data policies. We cannot change the incentive structure pushing commercial operators toward defense customers. Those are separate frameworks requiring separate political work.
What this Accord does is essentially try to build a global public EO infrastructure underneath the commercial EO layer, one that would make commercial fragility less consequential for the people who depend on EO.
Who Builds the EO Accords
The substance of an EO Accord is one question. Who builds it is another, and harder.
It helps to be exact about why Artemis worked, because the reason is not the one usually cited. Artemis ran on an engine with two parts: a single power with a reason to move first, and a rival to move against. The US wanted to set lunar norms before China did, and that rivalry did the heavy lifting for free: it gave the convener its motive and gave every signatory a side to pick.
An EO Accord has to give up both parts. No single power is positioned to convene it, and there is no rival to define it against — excluding China is a non-starter. So the real worry is whether this would work at all once those two core motivations are gone.
They can, but not for free, because EO's engine runs the opposite way. Rivalry gathers a coalition, but risk of losing access sends everyone off on their own. The rational response to watching your data access vanish or be restricted is to protect yourself — diversify suppliers, switch jurisdictions, go it alone — which is exactly what happened here: Planet's customers dispersed and Switzerland opted out while keeping access to the data and countries are all investing in their own national EO programs. Left alone, risk tends to fragment, not assemble.
So the coalition will not form by itself, the way Artemis's did. Someone has to convert scattered self-protection into shared commitment, deliberately. The two inversions below are what that takes: a different kind of convener, and a different posture toward the "rival".
From US-led to Coalition-Led
If not a single power, then who? The US is the clearest case of why the old answer may be gone: with shifting priorities for Earth science, it has become an actor whose EO budgets have become constrained and whose willingness to restrict access to data is on record.
Europe is the obvious alternative with Copernicus, the world's largest civilian EO program. But Europe alone is too narrow, and I don't mean just politically. The thing pulling signatories in is shared risk, which reads the same in New Delhi and Lagos as in Berlin. A convener credible only inside the West cannot gather a coalition whose whole point is breadth. It needs a partner with legitimacy across the North–South divide.
India is the natural candidate, through the G20 Open Satellite Data Partnership it put forward in 2025 and the G20 satellite that ISRO is launching next year. Neither leg is proven — a China-inclusive accord is a heavy lift, and from what I know, India's G20 move is yet to receive broad global traction. But, in any case, it is the most credible starting point on offer, not a finished coalition.
It does not have to be finished to begin. Artemis opened with eight signatories, not 67, so a credible founding group is enough to start. The catch is only that it has to be actively built. Risk will not hand anyone a coalition the way rivalry handed one to the US. Who convenes the founding and where the Accord eventually lives are separate questions, and only the first needs answering now. The convening question is still open.
From Deliberately Exclusive to Deliberately Inclusive
The other inversion follows from the same fact. Artemis could define itself against China because beating China was the goal, but an EO Accord cannot. China launches more EO satellites in a year than most countries combined and might be the source some users turn to when US providers restrict access. A framework with no Chinese path is a Western club, not an Accord — and a Western club reproduces the very fragmentation an Accord is meant to end.
The asymmetric structure of the EO Accord commitments is what makes inclusion workable rather than naive. China is unlikely to sign every commitment for every dataset, least of all the civilian/dual-use designation, where its position diverges sharply from the West. But it could plausibly commit to transparency, interoperability, provenance, and crisis sharing for specific designated datasets — in climate, ocean, and disaster response, where its own interest in participating in a stable global baseline could be real. Partial signature on the accessible commitments is not a failure of the model – it is the model. Asymmetric participation beats exclusion, because exclusion guarantees the very outcome the Accord exists to prevent.
Closing Note: So, Why Now?
I don't think the question is about whether some form of the EO Accord I am proposing here gets built. The more relevant question is whether it gets built deliberately, or assembled crisis by crisis: an ad-hoc fix after one restriction event, a bilateral agreement after a continuity gap due to the loss of one satellite (but with datasets that are not interoperable) – all of which resulting in a global EO system as a tangle of incompatible responses to yesterday's problems.
The users of EO have learned, recently and concretely, that their data sources can disappear overnight, that new missions can be rescoped or access to them can be limited by political decisions. Insurance, finance, agriculture, mining, energy, humanitarian operations, and even journalists are all diversifying their data sources in response. Basically, each of them is doing privately what an Accord would do collectively: picking reliable sources and building on the most stable data they can find. The longer that work happens individually, the harder it becomes to coordinate later – if we ever want to have a global baseline.
The cost of inaction compounds in the opposite direction. China is certainly launching EO capacity faster than anyone. The US seems to be stepping back from Earth science based on some specific decisions and budget proposals. Europe might be carrying the open data dependency alone, with dual-use investments competing for the same political attention and budgets. Sixty-plus countries are launching satellites that don't seem to be adding up to shared infrastructure. None of this gets easier by waiting.
The feasibility case might have gotten stronger too. The Artemis Accords have established that bilateral, principle-based frameworks can scale — 67 signatures in five years is no longer hypothesis, it is precedent. OceanEye has shown that a new European-led convening is possible. The weather coordination history shows the institutional artefacts can be constructed. While none of this guarantees an EO Accord (for reasons mentioned above), part of me feels like they demonstrate I am not utopian to think that one is feasible.
I want to spend the next few years working on this. The Accord this essay argues for will not get built by analysis alone. It needs people, especially the users who depend on EO data to be organised and loud enough that governments feel the cost of doing nothing. Those users exist, from Fortune 500 firms to impactful non-profits, but right now they are all scattered.
No organisation brings them together into a voice that governments hear, or want to hear. There is no global equivalent of an advocacy organization – like The Planetary Society for Earth observation. That is what I intend to build in the coming years. Building that might be the first step, and using it to push for the Accord in some form, is where the hard work is.
If you are thinking about this too, let's talk!
Until next time,
Aravind
