Posted by Alexander C. Kaufman

https://www.theatlantic.com/science/2026/04/stardust-solar-radiation-management/686664/?utm_source=feed

The world’s first major private geoengineering start-up must have known it’d have skeptics. The basic premise—dimming the sun to artificially cool the planet—has been called reckless by scientists and climate advocates; military analysts have said it has real security risks. (Don’t even get conspiracy theorists started.) Still, Stardust Solutions, an American-Israeli company planning to build a solar-reflecting system in the next few years, has enough fans that it raised $60 million in 2025. Now it’s trying to bring around the public and ease concerns over a technology for which its founders would prefer there wasn’t any conceivable global need.

For years, solar-geoengineering research has focused on sulfate aerosols that effectively mimic the impact of volcanic eruptions. Because the cooling effects of volcanoes are well studied, the problems that could arise from spraying sulfur into the atmosphere are also well understood: They include damage to the Earth’s ozone layer and acid rain. Stardust claims to have invented a unique particle that avoids those issues while matching the reflective powers of sulfur.

This week, Stardust is planning to release two documents that it hopes will help alleviate any fears over its next steps. (The company shared both with me in advance of their release.) The first is four pages of “guiding principles,” explaining why the company sees research into solar-radiation modification, or SRM, as essential. The second document outlines a 14-page framework for safely deploying it outdoors.

The documents are grand in their language and intricate in details. But neither document addresses the two key questions facing this initiative: What kind of particle does Stardust want to inject into the atmosphere, and why should a private company be trusted to control a planet-altering technology?

In its guiding principles, Stardust makes an argument for geoengineering that has quickly gained traction in the past couple of years among scientists and even environmentalists. Few advocate the use of geoengineering—right now at least. But as record heat waves turn deadly and polar ice melt threatens to set off uncontrollable natural cycles, a growing consensus is emerging that the technology needs to be at least studied. In that vein, the company stresses in the memo that “Stardust is pursuing R&D, not deployment.” Researching the technology now, it argues, is “essential for any future informed decision-making by governments on whether - and, if so, how - to deploy SRM.” The principles also promise that Stardust will be cautious and “committed to avoiding environmental harm” as it begins to test its technology this year, and that it will release its data, methods, and results, “including unfavorable ones,” in “a timely manner, following appropriate scientific validation.”

The framework in the second document focuses on testing. As a field, although admittedly a small one, geoengineering’s testing strategy has hitherto erred on the side of asking for forgiveness, not permission: In 2012, for instance, a U.S. businessman conducted an experiment that allegedly violated United Nations rules and involved dumping iron sulfate off Canada’s west coast in a bid to speed up carbon absorption. Ten years later, a start-up carried out a rogue experiment in Mexico, sending two weather balloons filled with sulfur dioxide and helium into the air. Stardust told me it has no outdoor tests planned and that it will not conduct any until relevant regulations are established; its framework proposes extensive laboratory testing and computer modeling before beginning small-scale testing that they will “tightly monitor,” eventually scaling up to dispersing the particles across a wider area.  

Still, “it’s just a framework,” without any specifics on research plans, Holly Jean Buck, a University at Buffalo researcher whose book After Geoengineering spells out potential benefits of this type of technology, told me. “Frameworks are everywhere in academia,” she said. Likewise, after I reviewed the details of the documents with her, she said that “the guiding principles just seem like generic things that the research community has been talking about.”

The most obvious details missing from the documents are any information about Stardust’s proprietary particle. “None of us knows what they are hoping to put into the stratosphere—for a profit,” Cynthia Scharf, a senior fellow at the Brussels-based Centre for Future Generations who studies geoengineering, told me. Rather, she said, the new documents “repeat blithe paeans to transparency, safety, and informed governance.”

Yanai Yedvab, Stardust’s CEO and a former nuclear scientist, told me that Stardust plans to unveil its particle “in the coming months.” And he insists the company wants to be regulated. I asked him what industry or company might compare to Stardust and its model—weapons production, for instance? Raytheon, after all, does not decide whether or where to fire Tomahawk cruise missiles; the use of massively destructive weapons is constrained, to some degree, by the types of international agreements that might govern geoengineering.

Yedvab offered an alternative: pharmaceutical companies, which are often funded by and work in partnership with governments, but which also work under strict regulation. “You’d never believe a pharmaceutical company telling you that a cure for cancer is safe,” he said. “You have, for this, the regulators and independent bodies.”

But right now, at least in the U.S., no obvious agency could act as the Food and Drug Administration for geoengineering. And Stardust is already trying to shape whatever regulatory environment it might face. In November, the environmental newswire E&E News reported that the company had hired the lobbying powerhouse Holland & Knight to begin “informing members of Congress about our work and the need for appropriate and robust oversight” of its research, Yedvab said in a statement at the time. (The lobbying shop had failed to publicly disclose its work for Stardust due to what it said was a clerical error.)

In its new guiding principles, too, Stardust forswears working with “entities likely to engage in irresponsible” activities with its technologies and argues that potential deployment “should only take place through credible and informed international decision-making.” Ultimately, the company envisions selling its services exclusively to national governments or international bodies, such as a United Nations agency. “We expect and hope that governments will come together in a few years to allow international decision making on SRM, much as they did with the Montreal Protocol,” Yedvab told me, referring to the 1987 treaty to phase out use of the chemicals causing a hole in the Earth’s ozone layer.

As a business strategy, it’s necessarily unusual. A global project to spray aerosols into the atmosphere could not have competition without unpredictable and potentially disastrous results. Who exactly might its customers be? Perhaps the U.S. and China, together, through some UN partnership; perhaps a middle power, such as Brazil or India, could buy Stardust’s intellectual property and engage the UN in some kind of global effort. Smaller countries already facing some of the worst effects of climate change, such as Ghana, recently started openly calling for access to their own geoengineering solutions. If governments don’t buy into Stardust’s technology, the company’s work could at the very least underscore how little progress governments themselves are making.

Certainly, private industry is producing some of the most powerful breakthroughs of our age: reusable commercial rockets, deep-ocean mining, chemistry that generates clean water from thin air, AI. Stardust could end up inventing a tool that wins over the public with transparent science and practices that prioritize public good. It could also be the West’s best bet to maintain its lead over a technology that could offer humanity one of its most profound instruments to reshape the world, given that China has so far made only modest research efforts.  

Once we start geoengineering, though, stopping may be difficult. Scientists and green groups have come around to the idea of researching these technologies more because of concerns over the failure of decarbonization efforts than a bullishness on geoengineering. Yedvab echoes the ideas of the emerging middle ground, that geoengineering might buy the world time to build enough nuclear reactors, solar panels, and carbon-capture projects to avoid reaching calamitous levels of heat. But he also told me that he sees geoengineering as a way to keep the standard of living we have now. This technology will “allow people to live more or less the life they aspire to live, the life they’re used to living,” he said. “We want to make sure our children will live in a world where they don’t suffer the horrors of climate change. While this is not the only tool and we should have the full portfolio, I think it’s a unique tool that enables you to hold both sides at once.”

For how long? If the world’s governments do use geoengineering to slow-walk a transition from fossil fuels—as critics of the technology fear—the world could be locked into decades or even centuries of spraying the atmosphere to maintain the habitability of our planet. All the while, we could end up paying private companies for that privilege. But the costs of not having some kind of insurance policy against runaway warming could be even higher.

https://www.theatlantic.com/science/2026/04/stardust-solar-radiation-management/686664/?utm_source=feed