One Startup Raised $60 Million to Dim the Sun. Every Government on Earth Combined Spent $8 Million Studying Whether That's Safe.
Private capital for stratospheric aerosol injection now outpaces global government research funding by at least 6:1. Companies are deploying atmospheric tests, writing their own safety standards, and lobbying Congress. No international governance framework exists.
Sixty million dollars. That is what Stardust Solutions raised in a single funding round, announced October 2025, to develop and deploy proprietary calcium carbonate particles that reflect sunlight back into space. For context, the most thorough tracking study of global solar geoengineering research funding, conducted by the Harvard Solar Geoengineering Research Program (now the Salata Institute), found total worldwide government and philanthropic spending on stratospheric aerosol injection (SAI) research was approximately $8 million per year as of 2018. Even generous estimates for 2024-2026 put the figure at $20-30 million per year once you include philanthropic contributions. One company's single fundraise exceeds two to seven years of global public research spending.
Stardust is not alone. Make Sunsets, a smaller startup, has completed 124 self-reported balloon launches releasing sulfur dioxide particles into the atmosphere, selling "cooling credits" as unregulated offsets. Mexico banned the company from operating within its borders in January 2023. In April 2025, the EPA issued a demand letter to Make Sunsets under Clean Air Act Section 114, requesting detailed information about its atmospheric releases. Neither country has enforcement mechanisms that extend to international airspace.
Meanwhile, the institution that tried to study this technology responsibly gave up. Harvard's SCoPEx, the world's most prominent academic SAI research program, abandoned its work in March 2024 after an advisory committee recommended against proceeding. Indigenous communities in Sweden had protested the proposed balloon test site. A program designed to produce exactly the kind of safety data governments need to regulate this technology shut down because the governance structures to authorize even a small-scale experiment did not exist.
Follow the Money: A 6:1 Ratio
Here is the math on private versus public funding for SAI. Stardust Solutions: $60 million (single round, October 2025). Make Sunsets: undisclosed total funding, but the company has paying customers for cooling credits and has conducted 124 deployments. Other private efforts exist but have not disclosed funding amounts publicly.
On the government side, the Bulletin of the Atomic Scientists reported that "the total amount of global funding supporting solar geoengineering research has been rather minimal," estimating roughly $1 million per year globally in 2008 and approximately $8 million per year by 2018. No follow-up tracking study has been published. If we assume a generous 2x increase since 2018, current global government and philanthropic spending might reach $16-20 million per year.
Stardust's single round: $60 million. Best estimate of annual global government research: $8-20 million. Even at the generous end, that is a 3:1 ratio. At the conservative end: 7.5:1. Factor in Make Sunsets and undisclosed private actors, and total private capital likely exceeds 10x annual government research spending. The entities that stand to profit from deployment are outspending everyone studying whether deployment is safe.
| Funding Source | Amount | Purpose | Accountability |
|---|---|---|---|
| Stardust Solutions (Series A) | $60M (one round) | Develop + deploy SAI particles | VC investors only |
| Make Sunsets (cumulative) | Undisclosed + credit revenue | SO₂ balloon releases | Customers buying credits |
| All governments globally (annual) | ~$8-20M/yr (estimated) | Research, modeling, assessment | Taxpayers, UNEP, national science agencies |
| Harvard SCoPEx (total lifetime) | ~$16M (2017-2024) | Small-scale field experiment | Advisory committee, academic review |
Governance Velocity: 18 Years and Counting
Every other dual-use technology of comparable consequence acquired at least a preliminary international governance framework within a decade of its emergence. Solar geoengineering has not. Here is the comparison, constructed from the year each technology became a concrete capability to the year binding or quasi-binding international governance appeared.
| Technology | Capability Year | First International Governance | Time to Framework |
|---|---|---|---|
| Nuclear weapons | 1945 | 1968 (NPT signed) | ~23 years |
| CFCs / Ozone depletion | 1985 (ozone hole confirmed) | 1987 (Montreal Protocol) | ~2 years |
| GMO crops | 1994 (Flavr Savr tomato) | 2003 (Cartagena Protocol in force) | ~9 years |
| CRISPR gene editing | 2012 (Doudna & Charpentier) | 2015 (Int'l Summit on Human Gene Editing) | ~3 years |
| Stratospheric aerosol injection | 2006 (Crutzen paper) | None (2026) | 20+ years, no framework |
Paul Crutzen's 2006 paper in Climatic Change is widely regarded as the moment SAI became a serious scientific proposal rather than fringe speculation. Twenty years later, the UNEP published a 2023 assessment calling for more research but proposed no binding framework. The closest thing to governance is a 2010 CBD (Convention on Biological Diversity) moratorium on geoengineering activities, which carries no enforcement mechanism and which the United States has not ratified.
Nuclear weapons got their framework in 23 years, but the stakes were existential and obvious: mushroom clouds focus the mind. CFCs moved in 2 years because the ozone hole was photographable and the solution (substitute refrigerants) was commercially viable. SAI governance is stalled because the technology is cheap, the effects are diffuse, the beneficiaries and victims may be on opposite sides of the planet, and no mushroom cloud forces the conversation.
One Company, Four Roles
Stardust Solutions occupies a position no other entity in the history of atmospheric science has held simultaneously. According to E&E News reporting from November 2025, the company is: (a) developing proprietary SAI particle technology, (b) writing its own 14-page safety principles document, released in early 2026, (c) conducting small-scale atmospheric tests at an undisclosed location, and (d) lobbying Congress through a hired lobbying firm, with filings listed under "Science & Technology" on OpenSecrets.
This is not analogous to pharmaceutical companies funding drug trials. Pharma operates within a regulatory structure where the FDA defines trial protocols, monitors adverse events, and holds veto power. No equivalent body exists for atmospheric intervention. Stardust's self-published safety principles document has no external review process, no enforcement mechanism, and no consequences for deviation. CEO Arjun Patel has described the company's work as a "temporary bridge" while emissions reductions scale. The bridge is being built while the building codes are being debated in a committee that does not exist.
New Research, No New Rules
Academic work continues even without Harvard SCoPEx. An April 2026 ArXiv preprint outlined safety and controllability requirements for SRM (solar radiation modification) systems. A Springer Nature paper addressed "termination shock": the rapid warming that would occur if aerosol injection stopped abruptly after years of continuous deployment. A separate modeling study found SAI could reduce malaria transmission risk in South Asia by lowering peak temperatures. Three data points. Zero governance structures.
Papers describing what safety standards should look like are not the same as institutions with the authority to enforce them. Termination shock research demonstrates a known risk of beginning deployment without guaranteed long-term commitment. Malaria benefit modeling shows that SAI would create winners and losers across different geographies — South Asia gains while sub-Saharan Africa may see altered monsoon patterns — exactly the kind of distributional question that requires international negotiation, not a 14-page white paper from a Boston startup.
Strongest Counterargument
Proponents of private SAI development make a reasonable case: if governments will not fund this research, someone has to. Climate change is accelerating. The IPCC says 1.5°C of warming above pre-industrial levels will likely be breached within this decade. Every year of delay means more irreversible damage to coral reefs, ice sheets, and agricultural systems. Private capital stepping into a void left by government inaction is not a market failure. It is a market responding to government paralysis. Stardust, whatever its motives, at least published safety principles. That is more transparency than most defense contractors offer for far more dangerous technologies. And the alternative to private experimentation is not careful public experimentation. It is no experimentation at all, while the atmosphere warms.
This argument deserves a serious response because it correctly identifies the core problem: government spending on SAI research is negligibly small relative to the stakes. If global warming causes $23 trillion in annual damages by 2050 (Nature, 2024), spending $8 million per year studying a potential mitigation tool is a rounding error of a rounding error. Stardust's $60 million, by this logic, represents rational capital allocation. But the counterargument proves too much. The same reasoning would justify any private actor with sufficient funding deploying any planetary-scale intervention, because the stakes are high and governments are slow. That is not a governance framework. It is the absence of one.
What We Did Not Prove
This analysis has important gaps. First, the $8 million per year global government figure comes from 2018 tracking data. No post-2018 tracking study exists. Actual current spending may be higher, though no evidence suggests it has closed the gap with private capital. Second, Stardust's atmospheric test details remain undisclosed. We cannot independently verify the scale, location, particle composition, or duration of their deployments. Third, Make Sunsets' 124 launches are self-reported, and the quantity of SO₂ released per launch is unknown. Fourth, lobbying influence is inherently difficult to quantify. OpenSecrets filings show that Stardust hired lobbyists, but the outcomes of those conversations are not public. Fifth, we compared governance timelines across technologies with very different political contexts. Nuclear governance was driven by Cold War dynamics. CFC governance was facilitated by available substitutes. Direct comparison requires the caveat that each technology's governance path was shaped by unique political forces, not just the passage of time.
What You Can Do
If you are a climate policy professional or work in science governance: push for your national science agency to fund SAI research publicly. The worst outcome is leaving all the data generation to companies with financial stakes in the results. The OSTP (Office of Science and Technology Policy) published a congressionally mandated report on SRM in June 2023, but no research program followed. Demand one.
If you are an investor evaluating climate startups: ask about governance risk. SAI companies operating in a regulatory vacuum today face the possibility of retroactive regulation, international bans, or liability claims from affected nations. Mexico already banned one operator. The regulatory risk premium for unilateral atmospheric intervention is not priced into current valuations.
If you are a citizen: Know that this conversation is happening without you. No public consultation process exists for atmospheric modification. Unlike nuclear energy (NRC licensing, environmental impact statements, public comment periods) or genetic modification (FDA advisory panels, USDA APHIS reviews), there is no forum where affected communities can object before particles enter the stratosphere. That absence is itself a policy choice.
The Bottom Line
We have the worst possible configuration for a planetary-scale technology: abundant private capital, minimal public research, no governance framework, and companies simultaneously developing the product, writing the safety rules, testing in the atmosphere, and lobbying the regulators. For every other technology of comparable scale, some combination of catastrophe, public outcry, or scientific consensus eventually produced governance. Solar geoengineering has not yet triggered any of those forcing functions, because its effects are slow, diffuse, and unequally distributed. By the time the forcing function arrives, the particles may already be overhead.