Stardust is tackling the wrong problem with the wrong structure

By David Keith | April 30, 2026

In late 2025, Stardust, a for-profit startup, announced it had raised $60 million in venture capital funding to develop a proprietary system for solar geoengineering. The company’s mission is to commercialize sunlight reflection methods (SRM) by deploying specially engineered, “chemically inert” (i.e., safe) particles into the stratosphere via custom aircraft. This is a system that Stardust claims could be operational by the start of the next decade.

I want to see a technically detailed plan for the early deployment of sunlight reflection using sulfur in the stratosphere. Such a plan should extend from materials handling and dispersal from aircraft, through monitoring hardware, data assimilation, and the feedback controls needed to ensure that deployment achieves a goal such as ramping up cooling at a rate of 0.1 C per decade—roughly halving current warming—while maintaining hemispheric balance.

I want this plan to be detailed and public. Such a plan would inform decisions about governance and deployment. A critical review of SRM requires such a plan.

The SRM research community is doing an inadequate job of producing such a plan. So, there is an unmet need.

Stardust has a strong technical team with systems engineering expertise that could address this need. Yet I don’t expect Stardust to meet this need, first because as an opaque, for-profit company, it has the wrong structure; and second, it’s focused on the wrong problem.

Daniele Visioni and I argued that Stardust has the wrong structure in our 2025 op-ed in MIT Technology Review; proprietary technology and profit motives are directly at odds with the transparency needed to legitimize SRM research. As we said, “trust will be the most important single ingredient in making these decisions. And trust is the one product for-profit innovation does not naturally manufacture.” That is why I want to see a public not-for-profit doing this work instead.

Here I argue that Stardust is focused on the wrong problem. Stardust seems focused on finding an alternative aerosol that is better than sulfates. Yet the direct risks of sulfur—as opposed to other SRM aerosols—are quantitatively small and well understood, whereas the solutions Stardust advances are all-but-certain to introduce new unknown unknowns, which might entail potentially large, unanticipated risks.

As I see it, the main attraction of Stardust rests on two assumptions1:

  1. Sulfur is a risky way to do SRM, so there’s a big benefit in finding particles that are less dangerous than sulfur.
  2. If an SAI aerosol was made from something non-toxic it would be safer than aerosol made from sulfate.

Let’s start with assumption #1. Suppose one could completely remove the direct health impacts from the sulfate aerosols injected into the stratosphere? Suppose there were a magic aerosol that causes zero environmental or health impacts as it makes its way down to the ground and into lungs and ecosystems. How much would this change the risk of SRM?

Not much.

While SRM will impact human health though changes in air quality, sulfuric acid’s direct contribution to that harm is surprisingly small.

Seb Eastham’s 2018 paper showed that the direct impacts of sulfate SRM on air quality are small compared to the indirect effects of the SRM’s cooling and other climate change on air quality.

The amount of aerosols in the air we breathe is most commonly measured as PM2.5, the total mass in micro-grams of all aerosol particles smaller than 2.5 µm in each cubic meter of air.

The epidemiological evidence that aerosol particles are unhealthy is very strong. Very roughly, this data shows that if you are exposed to an extra ~30 µg/m³ of PM2.5 your life expectancy decreases by about a year.

We expect that aerosols added to the stratosphere will make their way to the surface atmosphere and add to PM2.5 causing harm. Seb Eastham aimed to quantify this impact of SRM. To his surprise, he found that while stratospheric aerosols descending to the surface did contribute to PM2.5, that contribution was small compared to the increase in PM2.5 caused by cooling. Why did cooling make more PM2.5? Mostly because it increased the amount of nitric acid in aerosols, nitric acid that mostly comes from industrial emission of NOx. (For more details see section 3.2 of Seb’s paper.)

Now, a new study from Daniele Visioni’s group using more modern methods has confirmed that result.

Assumption #1 is therefore false. It would be great to eliminate the direct air quality impacts of sulfate SAI aerosol, but these impacts are quantitatively small compared to the indirect effects of SRM on air quality.

Context matters: air quality impacts do not seem to be seen as a leading concern about SRM. (See my list of concerns here). Moreover, the combined impacts of sulfate SRM on human health counting aerosols (PM2.5), ground level ozone, and increased UV caused by damage to stratospheric ozone, are at least 10X smaller than the health benefits of SRM in the form of reduced deaths from heat. See Tony Harding’s paper in PNAS, and our commentary.

Onto assumption #2. If an SAI aerosol was made from something organic and “non-toxic,” it would be safer than an aerosol made from sulfate. This statement seems so obviously true as to be inarguable. Drink a glass of sulfuric acid and you die; eat a tablespoon of flour and you’re fine. Of course, an SAI aerosol made from organic, edible food-safe material would be safer than an aerosol made of sulfuric acid, right?

Maybe not.

Materials that are safe to eat or drink can sometimes be dangerous if inhaled. Wheat flour is safe to eat (unless one is intolerant). Yet inhaling half micron wheat-flour aerosol particles is dangerous.

The health hazards of aerosol particles cannot be deduced from the biocompatibility or toxicity of the materials out of which they are made.

While there are well established protocols for testing the safety of materials ingested in food or drink, there are no similar protocols for testing aerosol safety, particularly not chronic exposure to low concentrations of aerosols.

In many environments, sulfate is a good fraction of total mass of PM2.5, so regulations have pragmatically focused on sulfates.

But that does not mean that it’s the sulfuric acid in the aerosols that is causing harm. Despite decades of studying the health effects of PM2.5, scientists are still uncertain about what specific components of particulate matter are causing the harm.

Evidence suggests sulfate itself may not be the root cause. Mice exposed to pure sulfate aerosols in the lab suffer comparatively low health impacts. Here’s a recent paper.

A leading hypothesis for why PM2.5 aerosols are dangerous is the so-called “hitchhiker hypothesis,” which is the idea that these aerosols pick up tiny amounts of organics or metals and transport them into the lungs. Once there, the hitchhiking organics or metals cause the harm, not sulfate itself2.

So, it’s plausible that some organic, food-safe aerosol could be more dangerous than sulfate if the main driver of toxicity is the hitchhiking organics or metals and not sulfate itself.

Moreover, a Stardust-engineered aerosol injected into the stratosphere will undergo complex chemical reactions as it descends though the atmosphere and reaches someone’s lungs. So, a test of toxicity of the pure stardust aerosol could not accurately predict the impact in the real world just as tests of the toxicity of pure sulfate aerosol do not correctly predict the health impacts of PM2.5 aerosols.

Assumption #2 is therefore unknown. An SAI aerosol made from something non-toxic might be safer or less safe than an aerosol made from sulfate.

What we can say for sure is that a novel aerosol would have more uncertain health impacts than sulfate. There would be more basis for concern about unknown unknowns.

Sulfur aerosols are not safe. We know they cause mortality and morbidity, but we can estimate the size of their impacts with some confidence because of many thousands of scientific studies over the last century. This means there is little basis for concern about unknown unknowns with sulfate.

 

Summary

Stardust has the wrong structure because trust is the one product for-profit innovation does not naturally manufacture.

Stardust is focused on the wrong problem because (a) the direct health impacts of SAI sulfate can be assessed with some confidence using the immense scientific literature on their health impacts; and (b) assessments using this literature find that the direct health impacts of SAI sulfate are a small contributor to the overall risks of SAI; and finally, (c) the health and environmental risks of a novel particle would be more uncertain than the risks of sulfates.

I am no naïve booster of sulfates—I published some of the early papers on non-sulfate aerosols3. In addition to reducing health impacts, designer aerosols offer the chance to reduce three important side effects of sulfate SRM: heating of the lower stratosphere, damage to the ozone layer, and alteration of the visual appearance of the sky. It’s worth researching better methods of SRM.

But these three impacts only grow large when a large amount of sulfur is used, as would be needed to cool Earth more than about 0.5 or 1 degree Celsius. If SRM is gradually ramped up (in my view the only sensible way to use it) it will take a long time—perhaps half a century—until SRM is cooling Earth enough for these side effects to be important. Thus, it will be a long time before there is a sound reason to use designer aerosols.

In the long run, some new aerosols will be demonstrated that are better than sulfate. But it is very hard to argue the humanity should start with anything other than sulfate, because of the depth of our knowledge about its hazards and the inherent uncertainty in any designer particle.

I am convinced there is an urgent need for end-to-end systems engineering for SRM. While I disagree with what Stardust is doing, my conversation with the Stardust founders left me with a strong sense that they have good intentions. I believe their work is motivated by a shared desire to do this systems engineering for SRM. I wonder if they over-focused on the ‘ideal particle’ concept because it’s easier to sell novel tech to investors than it would be to raise funds developing a sulfate aerosol system.

In any case, the rise of Stardust has convinced me of the need for a public non-profit entity that can do this systems engineering for SRM using sulfate aerosols.

 

Acknowledgements: I thank Ben Peltz and Dakota Gruener for editorial suggestions.

 

Notes

1 On April 2, 2026 Stardust released a “A proposal for the safety and controllability requirements that SRM systems should meet” on arXiv.

2 See, https://hsph.harvard.edu/news/metals-and-sulfate-in-air-pollution-mixture-may-contribute-most-to-asthma-hospitalizations/, or https://documents1.worldbank.org/curated/en/810141630705865331/pdf/Are-All-Air-Pollution-Particles-Equal-How-Constituents-and-Sources-of-Fine-Air-Pollution-Particles-PM-2-5-Affect-Health.pdf

3 https://davidkeith.earth/publication/photophoretic-levitation-of-engineered-aerosols-for-geoengineering/, https://davidkeith.earth/publication/solar-geoengineering-using-solid-aerosol-in-the-stratosphere/, https://davidkeith.earth/publication/stratospheric-solar-geoengineering-without-ozone-loss/.

 

MIT Technology Review: Why the For-Profit Race into Solar Geoengineering is Bad for Science and Public Trust

By David Keith and Daniele Visioni

Last week, an American-Israeli company that claims it’s developed proprietary technology to cool the planet announced it had raised $60 million, by far the largest known venture capital round to date for a solar geoengineering startup.

The company, Stardust, says the funding will enable it to develop a system that could be deployed by the start of the next decade, according to Heatmap, which broke the story.

As scientists who have worked on the science of solar geoengineering for decades, we have grown increasingly concerned about the emerging efforts to start and fund private companies to build and deploy technologies that could alter the climate of the planet. We also strongly dispute some of the technical claims that certain companies have made about their offerings.

Given the potential power of such tools, the public concerns about them, and the importance of using them responsibly, we argue that they should be studied, evaluated, and developed mainly through publicly coordinated and transparently funded science and engineering efforts. In addition, any decisions about whether or how they should be used should be made through multilateral government discussions, informed by the best available research on the promise and risks of such interventions—not the profit motives of companies or their investors.

The basic idea behind solar geoengineering, or what we now prefer to call sunlight reflection methods (SRM), is that humans might reduce climate change by making the Earth a bit more reflective, partially counteracting the warming caused by the accumulation of greenhouse gases.

There is strong evidence, based on years of climate modeling and analyses by researchers worldwide, that SRM—while not perfect—could significantly and rapidly reduce climate changes and avoid important climate risks. In particular, it could ease the impacts in hot countries that are struggling to adapt.

The goals of doing research into SRM can be diverse: identifying risks as well as finding better methods. But research won’t be useful unless it’s trusted, and trust depends on transparency. That means researchers must be eager to examine pros and cons, committed to following the evidence where it leads, and driven by a sense that research should serve public interests, not be locked up as intellectual property.

In recent years, a handful of for-profit startup companies have emerged that are striving to develop SRM technologies or already trying to market SRM services. That includes Make Sunsets, which sells “cooling credits” for releasing sulfur dioxide in the stratosphere. A new company, Sunscreen, which hasn’t yet been announced, intends to use aerosols in the lower atmosphere to achieve cooling over small areas, purportedly to help farmers or cities deal with extreme heat.

Our strong impression is that people in these companies are driven by the same concerns about climate change that move us in our research. We agree that more research, and more innovation, is needed. However, we do not think startups—which by definition must eventually make money to stay in business—can play a productive role in advancing research on SRM.

Many people already distrust the idea of engineering the atmosphere—at whichever scale—to address climate change, fearing negative side effects, inequitable impacts on different parts of the world, or the prospect that a world expecting such solutions will feel less pressure to address the root causes of climate change.

Adding business interests, profit motives, and rich investors into this situation just creates more cause for concern, complicating the ability of responsible scientists and engineers to carry out the work needed to advance our understanding.

The only way these startups will make money is if someone pays for their services, so there’s a reasonable fear that financial pressures could drive companies to lobby governments or other parties to use such tools. A decision that should be based on objective analysis of risks and benefits would instead be strongly influenced by financial interests and political connections.

The need to raise money or bring in revenue often drives companies to hype the potential or safety of their tools. Indeed, that’s what private companies need to do to attract investors, but it’s not how you build public trust—particularly when the science doesn’t support the claims.

Notably, Stardust says on its website that it has developed novel particles that can be injected into the atmosphere to reflect away more sunlight, asserting that they’re “chemically inert in the stratosphere, and safe for humans and ecosystems.” According to the company, “The particles naturally return to Earth’s surface over time and recycle safely back into the biosphere.”

But it’s nonsense for the company to claim they can make particles that are inert in the stratosphere. Even diamonds, which are extraordinarily nonreactive, would alter stratospheric chemistry. First of all, much of that chemistry depends on highly reactive radicals that react with any solid surface, and second, any particle may become coated by background sulfuric acid in the stratosphere. That could accelerate the loss of the protective ozone layer by spreading that existing sulfuric acid over a larger surface area.

(Stardust didn’t provide a response to an inquiry about the concerns raised in this piece.)

In materials presented to potential investors, which we’ve obtained a copy of, Stardust further claims its particles “improve” on sulfuric acid, which is the most studied material for SRM. But the point of using sulfate for such studies was never that it was perfect, but that its broader climatic and environmental impacts are well understood. That’s because sulfate is widespread on Earth, and there’s an immense body of scientific knowledge about the fate and risks of sulfur that reaches the stratosphere through volcanic eruptions or other means.

If there’s one great lesson of 20th-century environmental science, it’s how crucial it is to understand the ultimate fate of any new material introduced into the environment.

Chlorofluorocarbons and the pesticide DDT both offered safety advantages over competing technologies, but they both broke down into products that accumulated in the environment in unexpected places, causing enormous and unanticipated harms.

The environmental and climate impacts of sulfate aerosols have been studied in many thousands of scientific papers over a century, and this deep well of knowledge greatly reduces the chance of unknown unknowns.

Grandiose claims notwithstanding—and especially considering that Stardust hasn’t disclosed anything about its particles or research process—it would be very difficult to make a pragmatic, risk-informed decision to start SRM efforts with these particles instead of sulfate.

We don’t want to claim that every single answer lies in academia. We’d be fools to not be excited by profit-driven innovation in solar power, EVs, batteries, or other sustainable technologies. But the math for sunlight reflection is just different. Why?

Because the role of private industry was essential in improving the efficiency, driving down the costs, and increasing the market share of renewables and other forms of cleantech. When cost matters and we can easily evaluate the benefits of the product, then competitive, for-profit capitalism can work wonders.

But SRM is already technically feasible and inexpensive, with deployment costs that are negligible compared with the climate damage it averts.

The essential questions of whether or how to use it come down to far thornier societal issues: How can we best balance the risks and benefits? How can we ensure that it’s used in an equitable way? How do we make legitimate decisions about SRM on a planet with such sharp political divisions?

Trust will be the most important single ingredient in making these decisions. And trust is the one product for-profit innovation does not naturally manufacture.

Ultimately, we’re just two researchers. We can’t make investors in these startups do anything differently. Our request is that they think carefully, and beyond the logic of short-term profit. If they believe geoengineering is worth exploring, could it be that their support will make it harder, not easier, to do that?

David Keith is the professor of geophysical sciences at the University of Chicago and founding faculty director of the school’s Climate Systems Engineering Initiative. Daniele Visioni is an assistant professor of earth and atmospheric sciences at Cornell University and head of data for Reflective, a nonprofit that develops tools and provides funding to support solar geoengineering research.

Original post in MIT Technology Review

Preserving Earth’s Reflectivity: Supplementary Comments on “A Responsible Way to Cool the Planet”

By David Keith | September 21, 2025

Supplementary comments on “A responsible way to cool the planet” with Zeke Hausfather, an opinion essay in the NYT published 21 September 2025.

My guess is that a fair-minded understanding of the risks and benefits of deploying a small amount (example below) of sunlight reflection would yield surprisingly broad support for limited deployment.

Yet, any talk of limited deployment comes with one giant caveat: the fear that once started, it would be politically hard to stop. If there’s a big problem, it’s easy to stop. But if it starts and goes roughly to plan, then there will be a strong temptation to keep doing more, while weakening efforts to cut emissions or remove carbon.

How to set up political structures that avoid over deployment? There are clever proposals for voting rules that tie the right to vote in decisions about SRM to successful emissions cuts, but there are no easy answers.

Zeke and I propose a simple rule: limit use of SRM to maintaining Earth’s reflectivity against the decrease in reflectivity that will continue as pollution is cleaned up. This rule has two useful consequences. First, it ties use of SRM to clean up of pollution, and since this pollution is mostly from burning fossil fuels, this is roughly equivalent to tying use of SRM to cuts in fossil fuels; and second, it provides a non-arbitrary fixed upper limit on how much SRM can be deployed.

One can imagine this rule is simple enough to be useful in international negotiation over sunlight reflection.

Any such limit can obviously be overridden. It’s impossible to bind the hands of the future, and it would be unwise to do so, even if it were possible.

 

Notes:

Here’s what a small amount could look like: injecting sulfur into the stratosphere using high-flying aircraft. Starting in the early 2030s and gradually increasing over time, the injection could, after about 50 years, reflect enough sunlight to cool Earth by 0.5 C. That would lower projected 2080 temperatures from about 2.5 C above pre-industrial levels without SRM to around 2.0 C, while balancing the cooling across both hemispheres. This a bit larger than current cooling now caused by pollution, which is more like 0.4 C, but it’s similar magnitude. The human and environmental benefits of reducing temperatures by half a degree are enormous as demonstrated by the IPCC 1.5 report.

A strong temptation to keep doing more. This is geoengineering’s moral hazard, perhaps better called “mitigation deterrence”. It can arise from political exploitation or collective addition, as I describe here: https://davidkeith.earth/publication/toward-constructive-disagreement-about-geoengineering-a-shared-taxonomy-of-concerns-may-help/ . It can arise from political exploitation or collective addition as I describe here: https://davidkeith.earth/publication/toward-constructive-disagreement-about-geoengineering-a-shared-taxonomy-of-concerns-may-help/

Here are my previous NYT op-eds:

The New York Times: What’s the Least Bad Way to Cool the Planet?

The New York Times: Blocking the Sky to Save the Earth

The New York Times: A Responsible Way to Cool the Planet

Zeke Hausfather and 

Since the Industrial Revolution, burning coal and oil has filled the air with sulfur, shortening the lives of billions of people. In response, countries passed stringent air pollution laws requiring coal plants to scrub out sulfur and ships to switch to cleaner fuels. Global sulfur emissions have fallen some 40 percent since 2006. China alone has slashed them by about 70 percent.

We should celebrate cleaner air, but we also have to reckon with an unintended consequence. It turns out that by reflecting sunlight back into space, tiny sulfur particles protected Earth from about a third of the warming caused by human emissions of carbon dioxide. Now more of the underlying greenhouse gas warming is showing through, accelerating climate change. As The Economist recently put it, “If India chokes less, it will fry more.”

For some of us in the world of climate science, this raises a thorny question: Should we explore replacing the inadvertent cooling effects of sulfur with a cleaner, deliberate version?

Geoengineering the climate in this way is not a new idea; it was in the first U.S. high-level climate report that reached President Lyndon Johnson’s desk in 1965. While several options have been proposed recently, the most plausible way to make Earth more reflective is to use a small fleet of high-altitude aircraft to increase the amount of sulfuric acid droplets in the upper atmosphere. We know this can work; when volcanic eruptions put large amounts of sulfur in the upper atmosphere, such as Pinatubo in 1991, Earth was noticeably cooler for a few years.

We should take such an idea seriously because the costs of losing accidental sulfur cooling were made painfully evident this year when heat waves pushed temperatures above 120 degrees Fahrenheit in the Middle East and North Africa. If sunlight reflection could save lives and protect the environment, it is at least worth discussing.

Because sulfur is much more effective at cooling the planet when put into the upper atmosphere compared with what’s released into the lower atmosphere when we burn fossil fuels, we’d have to add far less of it. And for the same amount of cooling produced from burning fossil fuels, sulfur in the upper atmosphere would cause at least 100 times smaller health impacts.

Sunlight reflection is no panacea. Putting sulfur in the upper atmosphere will damage the ozone layer, allowing more ultraviolet radiation through. Even if it reduces deaths from heat and extreme weather, large-scale deployment could exacerbate climate change in some locations, perhaps by shifting rainfall patterns.

There is also a more fundamental limitation to sunlight reflection. It is effectively a Band-Aid that treats the symptoms of climate change but not the underlying disease of greenhouse gases. And unlike a skinned knee, the Earth does not heal from climate change on any time scale that matters for human societies. Warming from carbon dioxide is astonishingly persistent; much of what we emit today will warm the planet for many thousands of years to come.

Even if the world drives emissions down to zero, the planet wouldn’t cool down for millenniums. The only durable way to return to cooler temperatures is to remove the excess carbon we have already added, and removing enough to reverse even 0.1 degree Celsius of warming would cost tens of trillions of dollars.

This means that the real risk of geoengineering is not some Hollywood-style catastrophe, but complacency. A cheap way to delay the effects of warming risks undermining the need to rapidly reduce emissions, and going down that path would risk locking our children into a dependency where even stopping the process becomes too expensive to contemplate.

Given all this, we are not advocating deploying geoengineering today. But if policymakers decide that it is needed, a more modest approach would be to run a small, carefully scaled program that slightly increases the upper atmosphere’s reflectivity to compensate for the loss of cooling as sulfur pollution is eliminated.

The goal would not be to dial the Earth to some preferred temperature, nor to offset all greenhouse warming. It would be to keep the total cooling from sulfur roughly constant for a period of time, reducing near-term climate risk while decarbonization efforts continue.

As world leaders gather in New York for the U.N. General Assembly and Climate Week, any discussions of sunlight reflection should have a clear, enforceable commitment to never cool the Earth more than today’s current sulfur emissions do. And it should come with a clear off-ramp: As the world reaches net-zero emissions and scales up carbon removal technologies later this century, the program should end.

Pacing matters as much as limits. If society ever chooses to test this approach, it should start small and move slowly. Tying it to reductions in air pollution allows a slow ramp-up, resulting in increments that are imperceptible to most of us but visible to satellites and sensors. This should be coupled with regular checkpoints to assess side effects on regional rainfall, the atmosphere and ozone. The intent is to buy a modest, temporary buffer, not to start a new arm of climate control.

If we stay focused on the cure of reducing emissions and consider bounded and temporary sunlight reflection, we could preserve cleaner air, avoid a near-term temperature surge and not betray the generations to come, who will live with the consequences of our choices today.

Original post in The New York Times. 

Beyond Offsets: How to Build a Carbon Removal Industry

Carbon offsets aren’t enough. To truly tackle climate change, we need a global industry dedicated to pulling carbon out of the air and at massive scale. Join hosts John and Sonia inside the innovation race to scale carbon removal technologies, featuring insights from leading voices in the field.

They speak with Dr. David Keith, a pioneering climate scientist and founder of Carbon Engineering, who unpacks the technological, policy, and economic hurdles to direct air capture and other approaches. You’ll also hear from two recent XPRIZE Carbon Removal winners, Mike Kelland of Planetary Technologies and Jim Mann of UNDO about how their startups are using ocean alkalinity and enhanced rock weathering to permanently sequester CO₂, while also delivering benefits to farmers and marine ecosystems.

Together, they explore whether the world can build a scalable, measurable, and credible carbon removal industry – one capable of drawing down billions of tons of CO₂ annually.

Original post on RBC