Janos Pasztor was conflicted. Sitting in his home office in a village just outside Geneva, he stared into the screen of his computer, where a bizarre Zoom call was taking place. It was Jan. 31, 2024. The chief executive of an Israeli-U.S. startup, to whom Pasztor had only just been introduced, was telling him the company had developed a special reflective particle and the technology to release millions of tons of it high into the atmosphere. The intended effect: to dim the light of the sun across the world and throw global warming into reverse. The CEO wanted Pasztor, a former senior United Nations climate official, to help. The company called itself Stardust Solutions.
Pasztor, a deliberate and self-assured Hungarian with thick, arched eyebrows that give him the appearance of a mildly perturbed owl, was stunned by the seriousness of Stardust’s operation. He had long been expecting that some company would try this. But the emergence of a well-financed, highly credentialed group represented a shocking acceleration for a technology still largely confined to research papers, backyard debates and science fiction novels.
The Stardust CEO, Yanai Yedvab, was a nuclear physicist who was once deputy chief scientist at the Israeli Atomic Energy Commission, and he jumped straight to the point. He wanted Pasztor to advise him on how to build public credibility, which would be necessary to land the government contracts for sunlight reflection that the company and its investors were banking on. The CEO appeared keenly aware that Stardust had the potential for the kind of public image problems normally reserved for James Bond villains. Those challenges were likely not made easier by picking a company name that echoed Star Wars’ “Project Stardust” — the codename the bad guys in the Galactic Empire used for the Death Star, a weapon designed to destroy entire worlds.
For decades, scientists had theorized that lacing the atmosphere with a cloak of dust could temporarily reduce global warming. Few, however, had actually advocated researching the practice, and none could say how dangerously it might destabilize weather patterns, food supplies or global politics. Many scientists still warn it will take many years to know whether such technology would prove wise or disastrous. The terms for it — “solar geoengineering,” “stratospheric aerosol injection” or “solar radiation management” — sound deceptively anodyne. To most people, the idea of blotting out the sun still induces derision and disgust — a kind of planetary body horror.
If what Stardust was claiming on the Zoom with Pasztor was true, then a key threshold had already been crossed. Humanity had gained the power to turn down the sun, and barely anyone on the planet even knew. What’s more, that untested power was now effectively for sale. In a world of rising chaos, sci-fi-pilled billionaires and nationalist leaders, a private company offering the means to control the world’s temperature — with almost no international laws regarding the deployment of such technology — was a disturbing prospect, thought Pasztor.
But there was another consideration he couldn’t ignore: What if it was the only option left?
Despite huge advances in clean energy, greenhouse emissions were still rising. Even the upper limits on temperature that the international community agreed to in the 2015 Paris Agreement looked increasingly likely to be missed. In 2022, the U.N.’s climate science body said humanity was already locked into a series of terrible events: extreme heatwaves, uncontrollable wildfires and record-breaking floods and drought; the spread of tropical pestilence; the collapse of the world’s coral reef ecosystems. Those were expected to grow worse, depending on how quickly temperatures continued to tick relentlessly upward in the coming decades, driving mass wildlife extinctions, human conflicts and migration.
Unless, perhaps, solar geoengineering could intervene. “If the situation is as bad as it is, then we might actually need this technique,” the 70-year-old Pasztor said in September. He was seated at the dining table in his home, a half-hour train ride outside Geneva. Some sweet pastries he’d bought from the bakery across the square sat half-eaten on a plate. “If that is the case, then we have a moral imperative to study it — to understand it.”
Since 2016, Pasztor had run a think tank that used his sway as the former top climate adviser to then-U.N. Secretary General Ban Ki-moon to urge governments to take solar geoengineering seriously — as well as the need for global rules to manage it. But he had never advocated for its use. Strong rules would protect against several unpalatable but plausible scenarios, Pasztor said, including one in which “a billionaire who has very strong relationship with the government” — most likely the U.S. government — might just try it without any discussion or input from the global community.
That was why Stardust’s for-profit model was a matter of deep concern to Pasztor. Yedvab and his team had conducted research and built the 25-person company with virtually no oversight. Yedvab has repeatedly emphasized, in interviews with us and on Stardust’s website, that the company welcomes regulation and is taking steps to educate policymakers to that end. But currently, there is little that constrains it.
In addition to the lack of binding international regulations, there were also no robust national standards governing solar geoengineering in Israel, where the company had its laboratory, or in the U.S., where it was registered in Delaware, the state that’s home to a majority of the nation’s top companies. Stardust’s legal ties to the U.S., four finance experts said, hinted at its intentions to court investors from Wall Street and Silicon Valley and win U.S. government contracts. Pasztor feared its early backers, which included a venture capital firm led in part by former Canadian Prime Minister Stephen Harper and veterans of the Israeli security establishment, would have little incentive to allow scientific doubt or public fear to deter Stardust’s progress. (Neither the firm, Awz Ventures, nor Harper responded to requests for comment.)
Pasztor knew that working with such a company would strike some of his old friends and colleagues as an endorsement, even a betrayal. Another reason for his hesitation was that he was ready for a break. “I was going to be retired, really, and twiddle my thumbs and think, and maybe write,” he said.
A minute’s walk away from his apartment, the water of Lake Geneva splashed against a rocky shoreline where he and his wife, Christine, had spent the summer swimming. On clear days, they could float on their backs and look across to France and Mont Blanc, where one of Europe’s largest glaciers was rapidly withdrawing from existence.
Pasztor spent a couple of anxious days weighing what to do. Ultimately, he decided he had to find a way to work with Yedvab’s team, because they were clearly serious about bringing their technology to the world. “Like it or dislike it,” he said, “there is a Stardust, and they are developing an A-to-Z system to do stratospheric aerosol injection. It is happening.” He thought that by getting involved he might encourage the company to move out of the shadows — even a little — and that would be a win. Besides, he said, Yedvab had convinced him that he took the need for oversight and transparency very seriously. “They wanted to do right,” he said. Pasztor agreed to come on board as a consultant. His main task, he said, was to publish a report on Stardust’s governance and regulatory challenges and to lay out steps the company could take to open itself to public scrutiny.
Speaking in his home on a thundery Thursday afternoon, almost exactly a year after he published the report he was contracted to write and a few months after he ended his relationship with the company amicably, Pasztor was troubled. Apart from a link to his report on Stardust’s homepage, there was little public indication that they were taking his recommendations for transparency seriously. The company had not published a code of conduct it had agreed on with Pasztor and had told him it would release. The website itself was difficult to find while we were reporting this article; two cyber experts confirmed to us that it contained a line of code that hid it from search engines so that it could only be found using the link. (Yedvab said the coding that hid the website was unintentional and the company has since fixed it.) Some scientists in the geoengineering community were also complaining that Stardust remained secretive about the chemistry of its particle and its plans for releasing them.
Since Pasztor left, Stardust has been moving full speed ahead. In October, the company announced a fundraising milestone: It had secured $60 million from funds linked to, among others, Silicon Valley luminaries and an Italian industrial dynasty, bringing the total amount raised by the two-year-old startup to $75 million. It was a stunning development, far larger than any previous investment in solar geoengineering, and it placed Stardust way out in front of the few other groups looking to turn it from a theory into a business. “Don’t want to say only serious game in town,” said Gernot Wagner, climate economist at Columbia Business School and author of Geoengineering: The Gamble. “But yeah.”
Behind the scenes, there’s more evidence of a dizzyingly fast timeline. Confidential documents shared with POLITICO Magazine, public records and interviews with more than three dozen scientists, investors, legal experts and others familiar with the company reveal an organization advancing rapidly to the brink of being able to press “go” on its planet-cooling plans. Meanwhile, Stardust is seeking U.S. government contracts and quietly building an influence machine in Washington to lobby lawmakers and officials in the Trump administration on the need for a regulatory framework that it says is necessary to gain public approval for full-scale deployment.
The growing gap between Stardust’s progress and the public’s knowledge of the company has unsettled many scientists. “The way they’re going about it is absolutely the most toxic, worst way,” said Daniele Visioni, a climate scientist at Cornell University who spent half a day speaking with the company at the beginning of summer 2024 about their plans after they asked him for advice. Visioni could not get answers to many of his questions after refusing to sign a non-disclosure agreement. “There is nothing that makes people lose trust more than saying, ‘Oh this is all secret. But trust us.’”
As a private company whose technology is subject to little regulation, Stardust and its backers have no legal obligations to adhere to strenuous safety principles or to submit themselves to the public view. In fact, being totally transparent would be extremely rare for a startup. But when the stakes are so high, many critics of the company have said what’s normal for a startup isn’t really the point. For Pasztor, and a widening circle of researchers and government officials, Stardust’s perceived failures to be transparent about its work and technology have triggered a larger conversation about what kind of international governance framework will be needed to regulate a new generation of climate technologies.
The matter could hardly be more urgent; sunlight-reflecting companies such as Stardust could be capable of significantly altering the climate of the entire world in only a few years — and potentially wreaking unforeseeable environmental and geopolitical turmoil.
This was why Pasztor wanted to participate in this article.
He clearly had growing doubts about Stardust’s intentions. Was it possible Stardust had simply used him to put a respectable face on their operation without taking his concerns seriously?
Pasztor shook his head. “I don’t want to be a totally naive idiot,” he said. His apartment was filled with books and artwork collected during a lifetime of international postings. For the first time during a five-hour conversation, he appeared unsure of himself. His name is the only one that appears on Stardust’s website.
“I mean it worried me, and I thought about it quite a bit. Are these guys asking me just to use me, to have my name there and then they can do whatever they want? And I didn’t get that feeling during the time I worked with them. But that danger is there,” he said. Pasztor paused, then the moment of doubt was gone. “And what can I do about it? I could have said: ‘OK forget it, I don’t want to deal with you, period.’ That’s not what I did.”
Heavens and Earth
Stardust’s plan wasn’t all futuristic. The company’s technology is based on a process nearly as old as the Earth itself.
In June 1783, a 16-mile volcanic fissure blew open the southern side of Iceland. “First the ground swelled up with tremendous howling, then suddenly a cry shattered it into pieces … exposing [the Earth’s] guts, like an animal tearing apart its prey,” recalled Jón Steingrímsson, a local pastor. He survived the ordeal to write an account that was published long after his death. It remains one of his country’s earliest and most important autobiographical works.
For the next eight months, lava spewed from the earth. The sun was hidden by ash and smoke. One in five Icelanders died in the aftermath. Steingrímsson himself only escaped by good luck — or perhaps something more divine. One terrible day, a great wave of lava swept toward his church and village. The pastor gathered his congregation and delivered a sermon of such mighty power and devotion that, it was said, God himself diverted the course of the fire.
As the eruption went on, sunlight grew dimmer far beyond Iceland’s shores. The Laki eruption, as it would come to be known, had sent 122 million metric tons of sulfur into the sky. Much of it would reach the stratosphere, the placid layer of the atmosphere that begins between 4 and 12 miles above the Earth’s surface. Those particles drifted on barometric currents around the Northern Hemisphere, wreaking havoc on the world’s weather. China and Egypt were hit by drought, then famine. In North America and Europe, winter was exceptionally brutal. In February, the Mississippi River froze down to New Orleans. Ice floes were seen bobbing into the subtropical waters of the Gulf of Mexico.
Benjamin Franklin, who was then serving as ambassador to France from the newly independent government in Washington, wrote from a frigid Paris that the sun’s rays were “rendered so faint … that when collected in the focus of a burning glass they would scarce kindle brown paper.” In a striking insight, Franklin suggested the bleak weather was related to reports from Iceland of a “vast quantity of smoke” belching from the land.
Scientists now know that sulfur particles from volcanic eruptions hang in the atmosphere obscuring the sun like an umbrella of dust. Really big booms can have a noticeable cooling effect on the planet that lasts a year or more. In 2019, researchers concluded that Franklin’s hunch about Laki was right.
Stardust claims to have developed a system that can replicate and maintain the global cooling effects of a volcanic eruption, without all the lava and sulfur. The mechanics would be quite simple. Stardust envisages a fleet of around 100 planes — to begin with — flying into the stratosphere to deliver payloads of their particles, landing to reload, then immediately taking off again to repeat, continuously, every flight a tiny volcanic cough. Researchers, including Visioni, found last year that the most efficient way to achieve a steady, uniform decline in the global temperature would be to spread the particles from the regions just north and south of the tropics. That means launching from at least two places, for example Florida and southern Brazil. The particles would then spread around the globe producing a gradual, uniform decline in the global temperature, before eventually dropping out of the sky after around a year, according to Stardust, and needing to be replaced. The particles would reflect a very small proportion of sunlight back into space, but enough to cool the Earth.
Given the mind-blowing dollar figures that come attached to climate change, solar geoengineering would be comparatively cheap. “The cost is small,” said Douglas MacMartin, an aerospace engineer at Cornell University. It would be something along the lines of “tens of billions of dollars per year.” That’s what Stardust hopes will make it appealing to the governments it is courting as potential customers. For a country like the U.S., climate and weather-related disasters are growing ever more costly. In 2024, the bill was $182.7 billion. Faced with this, a few tens of billions could look like a bargain.
That the global temperature would drop is not in question. Britain’s Royal Society, the august scientific club that counts Isaac Newton, Charles Darwin and (him again) Franklin among its past members, said in a report issued in early November that there was little doubt it would be effective. They did not endorse its use, but said that, given the growing interest in this field, there was good reason to be better informed about the side effects.
Solar geoengineering is hardly a cure — more Wegovy than smallpox vaccine. But that doesn’t mean it can’t have broad benefits when weighed against deleterious climate change, according to Ben Kravitz, a professor of earth and atmospheric sciences at Indiana University who has closely studied the potential effects of solar geoengineering. “There would be some winners and some losers. But in general, some amount of … stratospheric aerosol injection would likely benefit a whole lot of people, probably most people,” he said.
Other scientists are far more cautious. The Royal Society report listed a range of potential negative side effects that climate models had displayed, including drought in sub-Saharan Africa. In accompanying documents, it also warned of more intense hurricanes in the North Atlantic and winter droughts in the Mediterranean. But the picture remains partial, meaning there is no way yet to have an informed debate over how useful or not solar geoengineering could be.
Some researchers still don’t think the field should be given time and resources. Speaking to reporters in September, Martin Siegert, professor of geosciences at the University of Exeter, said large-scale solar geoengineering would mean severing the natural link between the temperature of the planet and the concentration of greenhouse gases in the atmosphere. It was, he said, “an extremely profound and challenging thing for me to accept that that would be a good way forward for us when decarbonizing, of course, is such an obvious and necessary priority.” And in fact, carbon will continue to build up as long as the world continues to burn oil, gas and coal — even if sunlight-reflecting technology is put to use. The additional carbon would require more and more cooling to counteract — good for a business like Stardust. This would do nothing to prevent ocean acidification, localized air pollution and other harms from drilling, mining, processing and burning fossil fuels.
And then there’s the problem of trying to stop. Because an abrupt end to geoengineering, with all the carbon still in the atmosphere, would cause the temperature to soar suddenly upward with unknown, but likely disastrous, effects. This extremely worrying proposition is referred to by scientists as “termination shock.” This raises the risk, Pasztor said, of “extortion” by the companies or governments with control over geoengineering. In response to a question about this concern, Yedvab wrote: “Governments have established mechanisms to prevent misuse of technologies, such as over-pricing, by their contractors, and we are confident the same would apply here.”
Once the technology is deployed, the entire world would be dependent on it for however long it takes to reduce the trillion or more tons of excess carbon dioxide in the atmosphere to a safe level — a process that would involve untold armies of industrial carbon suckers, mashed up carbon-bonding basalt spread across the fields around the world, massive reforestation and a host of other novel efforts. All of which have their own downsides. Scientists have estimated that it could take a century or more. If humanity still can’t kick its fossil fuel addiction and successfully scrub carbon from the sky, scientists say the world would need to stay on the Stardust treatment plan indefinitely.
Citing all those reasons, almost 600 academics signed a 2022 call for a “solar geoengineering non-use agreement.” But the international community has taken little action to restrict the idea. One U.N. body, the Convention on Biological Diversity, has urged against geoenginneering “until there is an adequate scientific basis.” The U.S. is not a signatory. Last year, a Swiss proposal to set up an expert group to study the “risks and opportunities” of the field failed to advance at the U.N. Environment Assembly in Nairobi, Kenya. It was the second time the measure had come up short, with nations deeply divided over how to confront the issue. The matter is not on the agenda of the U.N.’s COP30 climate talks, currently being held in Brazil.
These disagreements foreshadow another major downside of the technology: its potential to make heavy geopolitical weather. In a 2023 report, the Defense Science Board, a panel of scientists convened by the Pentagon, dismissed the oft-discussed scenario of a rogue actor — a state or extremely wealthy person — covertly deploying solar geoengineering. Flying hundreds of planes into the stratosphere required a major country’s backing, they said; otherwise countries with supreme air power could easily stop it. But the Board did call for the U.S. to establish monitoring capabilities for detecting unusual particles in the stratosphere — a recommendation the Biden administration began to implement. The goal of the monitoring system, the Board said, was to ensure that “an adversary cannot be positioned for strategic surprise in executing unilateral climate interaction action.”
What the U.S. security community is primarily concerned about, said Erin Sikorsky, the director of the Center for Climate and Security, a research center based in Washington, is preventing geoengineering from being taken up by China or, perhaps, Russia — countries the U.S. can’t always rein in.
No experts said they were aware of any advanced geoengineering program in China. But it only took Stardust, with a couple dozen employees and $15 million in seed funding, two years to come up with its system.
The capacity for “misperception” and “disinformation” would also be huge, said Sikorsky. For example, if one country were suddenly struck by a punishing drought, it might blame — rightly or wrongly — a neighbor conducting geoengineering. “There’s a lot of escalation dynamics, whether it’s India and Pakistan, over water and climate issues between those two countries, [or] China and the U.S,” she said. In 2023, Sen. Marco Rubio, now U.S. secretary of State, referred to geoengineering as “the stuff that people are worried is going to start a war.”
Then there are ethical — even cosmological — questions. “Do we have the right to do this?” said Cynthia Scharf, a former colleague of Pasztor’s at the U.N. who later worked with him on geoengineering governance issues. “Thinking about future generations, do we have the right not to?”
The parasol Stardust wants to build in the sky would, at times, be visually beautiful. In the evening, the aerosols could extend the colors of the sunset, creating dramatic whole-sky light shows. It would be humanity’s largest ever, entirely deliberate attempt to change the fabric of nature. It’s a little on the nose, perhaps, that Stardust’s first address in a science park on the outskirts of Tel Aviv was Oppenheimer 4. The street is named after a botanist and not the father of nuclear weapons, but just like the invention of the atomic bomb, solar geoengineering would give rise to questions that have troubled scientists, theologians and storytellers through time.
Scharf mentioned Greek myths “like Pandora’s box or Icarus flying too close to the sun.” These ancient stories ask, she said, “What is humanity’s place in the greater cosmos, in the greater natural system — or creation if you’re a religious person?”
Given the nature of these dilemmas, it feels fitting that today’s debate over geoengineering should, in some small way, be informed by the two-and-a-half-century old eyewitness accounts of an Icelandic eruption by Franklin and Steingrímsson: a scientist and a priest.
A crisis of faith
It would be hard to find anyone more firmly part of the climate establishment than Pasztor. He has been part of every major negotiation and deal from the very beginning of international efforts in the early 1990s. By the end of that decade, when he was working for the U.N. climate change office, the annual talks had delivered a global deal on reducing greenhouse gas pollution called the Kyoto Protocol. It felt to Pasztor like the environmental movement was winning.
“We were happy, we achieved things, you know, we were delivering stuff. We had a treaty,” he said.
But, as the years went on, global temperatures continued to climb. The world remained hooked on fossil fuels, and the political will to kick the habit began to fray.
Pasztor can remember the exact moment his faith in global politics to deliver a solution dimmed. In July 2015, when he was serving as senior climate adviser to U.N. Secretary General Ban, he was in a meeting in Luxembourg. Preparations were being made for a U.N. climate conference in Paris later that year that was supposed to deliver the deal to solve climate change. But then-U.S. Secretary of State John Kerry arrived with an ultimatum. There was zero chance, Kerry said, the Americans would agree to legally binding limits on their pollution. Pasztor was floored. There was no way, in his opinion, that a voluntary system could deliver carbon emission reductions at the speed needed. “To me that was a big, big, big shock,” Pasztor said.
“Maybe I was just naive, but I always felt that actually the climate problem is solvable,” he said — after all, “the technologies and everything is there.” But those solutions needed political backing and countries working together, despite their differences, to overcome entrenched opposition from the rich and powerful fossil fuel industry.
Ten years on, Pasztor believes he has been vindicated; the Paris climate agreement did get countries to set plans for winding down their emissions and has likely staved off the very worst of climate catastrophes. But the voluntary pledges have fallen short, and the world’s temperature is still rising at a rate far beyond what scientists consider safe.
“I mean, we can see it’s not working,” Pasztor said at his home in Switzerland.
This downbeat assessment is by no means the consensus among climate experts. Some argue that thanks to the investment unleashed by the Paris deal the clean energy market now follows its own, unstoppable economic logic, making government pledges less important. Kerry’s spokesperson Matt Summers said: “The world is in a more sustainable place because of Paris. These are facts. Anything else is an exercise in magical thinking.” Still, most experts agree that the future of climate change is still defined by varying degrees of bad.
Pasztor had built his life on faith in international institutions to address the kinds of problems — like climate change — that were too big for any one country to solve on its own. But at that moment in 2015, he realized the international system could not stand up to the competing demands of domestic politics. A new strain of nationalism was on the rise — one that would soon bring Donald Trump to power. In June 2017, Trump announced he would withdraw the U.S. from the Paris deal, a move he repeated this year.
At the same time, solar geoengineering was moving from the fringes of scientific discourse to an area of legitimate academic inquiry. In 2012 and 2017 respectively, the University of Washington and Harvard began research programs in the field. Both schools were eventually forced to halt outdoor experiments that attracted the ire of local residents and elected leaders.
In 2012, Scharf had stopped Pasztor in a corridor in the secretary general’s office in New York and told him she had begun reading about geoengineering and was freaked out by the complete lack of regulatory safeguards. With no relish, she told him: “I bet you that at some point we’re gonna come to this.” Pasztor was aware of the concept, but had never considered it would be needed. He dismissed Scharf’s concerns. “Janos thought I was being very pessimistic,” she recalled.
But after Paris, his thinking shifted. In 2016, just as Pasztor was retiring from the U.N., he received a call from Irene Krarup, director of the V. Kann Rasmussen Foundation, a Danish philanthropic group. She asked whether he would be interested in launching a think tank to study geoengineering — and, in particular, to persuade governments to take seriously the need for global rules to manage it.
“She saw what was coming,” said Pasztor. “It took me 30 seconds to say ‘yes.’” (Krarup confirmed the conversation took place.)
The work he did with the Carnegie Climate Governance Initiative, which disbanded in 2023, led eventually to the call from Stardust.
Pasztor’s initial hesitation about working with Stardust wasn’t only because the startup aimed to profit from geoengineering. The fact that Stardust was partially an Israeli company was also a big source of doubt. He had been outraged by Israel’s war on Hamas that had killed thousands of civilians in Gaza. He believed Israel had violated international humanitarian laws. (The Israeli government recently rejected a U.N. report that concluded the country had committed genocidal acts in the war.) Ultimately, his belief in the need for geoengineering regulation won out, and he began working with Stardust not long after the 2024 call. He donated his $27,000 fee to the United Nations Relief and Works Agency for Palestine.Pasztor says he spent several months that year conducting interviews with Stardust’s top brass. He published an 18-page report in September 2024 on his LinkedIn account. (Pasztor published the report before there was a company website, and the site later linked to it.) The report laid out what he had learned about the company’s inner workings — which was relatively little.
Stardust and its investors were betting that a for-profit startup could move faster and without the social and bureaucratic constraints of a university research program. But Stardust’s speed had come at the expense of transparency, Pasztor told the company in his report, and that was a vital component for building public trust.
Pasztor’s report made a series of recommendations aimed at opening Stardust to scrutiny. Most notably, he urged Stardust to publicly declare a moratorium on deploying its technology until there was adequate scientific research to understand the potential risks and benefits. Pasztor also asked them to consider handing control of the technology to a philanthropy. In a response to POLITICO Magazine, Yedvab said of the moratorium, “In general, it is not for companies to tell policymakers what are the mechanisms they should adopt in this process.” He also said: “We welcome any contribution of philanthropy in this partnership. It should be stated that up until now, relying solely on currently available philanthropic resources would not have enabled us to conduct a sunlight reflection R&D program at the scale and seriousness that is required.”
Before Pasztor published the report, he says he ran through it with Stardust, testing to see which of his proposals Yedvab would object to. He was encouraged by the response. The only significant dispute was over his requirement the company publicly announce small-scale tests before they took place. He said to his wife Christine, who was more skeptical about geoengineering than he was, “This is crazy, because they are basically accepting everything I’m suggesting to them.”
In response to a question about this characterization, Yedvab said: “Janos has helped significantly in shaping our thinking.”
Together Stardust and Pasztor also developed a voluntary code of conduct, which included a declaration now repeated on its website that “sunlight reflection deployment should be conducted under established governance, guided by governments and authorized bodies.” Though Pasztor admits that these commitments would only last as long as Stardust chooses to adhere to them. Yedvab confirmed that Stardust agreed to publish a voluntary code of conduct but said the company is still working on it.
Stardust’s bare-bones website does gesture at many of the themes on which they agreed in the code. The website commits to government-led deployment and declines to cooperate with “entities likely to engage in irresponsible deployment or misuse of our work.” It also commits to publishing the results of its research and backs “comprehensive regulation.” It does not have details of its investors, which Pasztor’s report recommended, nor does it include Pasztor’s call for a moratorium.
So far Stardust has not opened its technology up to public scientific scrutiny. Stardust’s ongoing secrecy has left solar geoengineering experts confused and frustrated. Stardust has tried to consult widely among those working on sunlight reflection science. But researchers are typically asked to sign a non-disclosure agreement to protect the company’s intellectual property. While not an unusual move for a company with a potentially valuable invention, this is another source of tension with specialists in the field who see such contracts as antithetical to scientific freedom and peer review, especially for a technology that would affect every person on the planet.
(Pasztor said he did sign an NDA. It “was the first thing Stardust wanted to tackle,” he said. Though he described the whole exercise as a “waste of time and energy” because he deliberately never asked them about the scientific inventions covered by the document, only about their approach to governance and transparency, which was the subject of his report.)
In response to a question about NDAs, Yedvab referred us to a section of the website that pledges that in the beginning of 2026, the company will start publishing “key findings of our research together with our esteemed academic collaborators in peer-reviewed scientific literature.” He said most scientists Stardust had approached agreed to sign the NDA.
Stardust claims to have solved many technical and safety challenges, especially related to the environmental impacts of the particle, which they say would not harm nature or people. But researchers say the company’s current lack of transparency makes it impossible to trust. “There’s absolutely no reason to believe anything they say without them providing evidence of it,” the Cornell aerospace expert MacMartin said flatly.
Visioni and several other scientists questioned whether this startup could so badly damage the image of the field that it would set back philanthropic- or government-funded research by decades.
Another researcher and fellow climate startup veteran, University of Chicago’s David Keith, said he had refused to counsel Yedvab’s team. Keith founded the carbon dioxide removal company Carbon Engineering in 2009 and sold the firm to Occidental Petroleum 14 years later for more than $1 billion. But he has long opposed efforts to commercialize solar geoengineering because — unlike carbon removal where the potential negative impacts are local — spraying particles into the stratosphere would have planetary consequences. He believes the public can’t truly weigh those risks if it’s asked to place its trust in an inherently opaque corporation.
“My only advice was that they not be a for-profit company,” said Keith, who previously helped create Harvard’s solar geoengineering program. “I think that they are foolish, and I told them that.”
To Pasztor, dismissing Stardust is not an option; its weaknesses only highlight a larger gap in need of addressing. “Society has to wake up and figure out whether you want to ban them or you want to create governance frameworks within which they can operate,” he said. Despite his disappointments, he still sees international cooperation as a vital guardian.
‘This thing’s going mainstream’
Just before the visit with Pasztor in Switzerland, a document was sent to POLITICO Magazine. It was a “pitch deck” Stardust had shared with potential investors as it sought a second round of funding. The slideshow was marked “proprietary and confidential” and dated 2023. If any doubts remained about how seriously Stardust should be taken, the document, reported here for the first time, dispelled them. The gap between what it contained and what the company had said publicly was very large. Yedvab confirmed the document was authentic but said the company’s latest decks were significantly updated and that it’s thinking was now “a bit more nuanced.” He declined to share the new materials with us.
The document laid out a startlingly ambitious roadmap for Stardust’s plans. The timeline, for one thing, was jarringly short and raised serious questions about how there could be time for research and regulation, when, for example, it aimed to begin a “gradual temperature reduction demonstration” in 2027.
The presentation also included revenue projections and a series of opportunities for venture capitalists to recoup their investments. Stardust planned to sign “government contracts,” said a slide with the company’s logo next to an American flag, and consider a “potential acquisition” by 2028.
By 2030, the deck foresaw a “large-scale demonstration” of Stardust’s system. At that point, the company claimed it would already be bringing in $200 million per year from its government contracts and eyeing an initial public offering, if it hadn’t been sold already.
“Global full-scale deployment” could begin by 2035, the timeline said. This brief development-to-deployment arc coincided neatly with the investment cycles of many venture capital firms. VCs generally raise money from investors, buy stakes in risky but promising startups and then seek to deliver outsized returns within a dozen years, when some of the companies they invested in go public or are sold to larger firms.
The annual cost of the initial stage of global cooling, according to the pitch deck, would be around $20 billion per year, in line with previous cost estimates for solar geoengineering. Even if the U.S. government were to entirely foot the bill, that’s still only around a fifth of the annual amount analysts predicted Joe Biden’s Inflation Reduction Act would have spent cutting emissions. The expected payoff for Stardust from going stratospheric would be $1.5 billion in annual revenue. Elsewhere in the 2023 pitch deck were pictures of the stardust itself, a fine white powder heaped up in a manner that recalled the scene in Scarface, where Al Pacino’s paranoid drug lord sits at a desk behind a small mountain of cocaine.
Another page told potential investors Stardust had already run low-altitude experiments using “test particles.” As proof, it provided photographs of a single-engine-light aircraft flying low over an airfield runway. A puffy white trail stretches out behind it. Beside the airstrip, a man leans over a laptop, and beside him, two others take a selfie. Using a registration code printed on the underside of the wing, we were able to track down the plane’s owner, Roy Ben Anat, an Israeli aerial enthusiast who built his aircraft himself. “The guys were great and it was also fun,” he said when contacted through a private message on LinkedIn.
Stardust’s tests, details of which have never been revealed to the public, did not break any laws in Israel — providing the substance was, as Stardust would later clarify, not an environmental pollutant — according to Alon Tal, a professor of public policy at Tel Aviv University. However, Tal added, the fact that experiments geared toward geoengineering could happen outdoors in Israel with no public notice or discussion was, evidence of, legally speaking, “significant lacunae.”
The deck also suggested that Stardust had the brains and backing to pull off its ambitious plan. The roster of employees and advisers featured in the document included a chemist, an aerospace engineer, two nuclear physicists, two particle physicists, two nanotech scientists and a pair of former officials from the Israeli prime minister’s office. The document also claimed an impressive list of collaborations with leading universities, though only one of these could be confirmed. Stardust worked with renowned aerosol expert Ruth Signorell at the Swiss university ETH Zurich. (Signorell did not respond to requests for comment. ETH Zurich confirmed through spokesperson Franziska Schmid that they had a contract with Stardust through to spring 2027 to research “mineral particles.”)
The pitch deck isn’t the only indication the company is moving fast. Stardust’s ground game in Washington is equally advanced. It has contracted with lawyers, lobbyists and advisers who previously served in the White House, Congress and the Pentagon. That includes Red Duke Strategies’ Sherri Goodman, a former deputy undersecretary of Defense known as a heavyweight on climate and security issues. It was Goodman who initially connected Stardust with Pasztor and federal researchers. Goodman said she couldn’t recall when she started working for Stardust but said the relationship ended before Trump returned to office.
The startup continues to work with WestExec Advisors, a consulting firm founded by former Secretary of State Antony Blinken and other veterans of the Barack Obama and Biden administrations.
The firm is helping Stardust to shape potential U.S. and international regulations around solar geoengineering, according to Columbia Law School professor Matthew Waxman, a former national security official during the George W. Bush administration who is working for WestExec on behalf of Stardust. That process of establishing guardrails for potentially using the technology, Yedvab said, is likely to stretch across multiple administrations.
Solar geoengineering is “a solution that needs to have an horizon of decades. It’s not a solution for, you know, a single president,” he said, avoiding mention of Trump, who has fought domestic and global efforts to limit climate change.
Yedvab said the work with these Washington firms was largely about regulation. “Since there is no playbook for these requirements or formal governance and regulatory framework, the only way to establish this list of requirements is by consulting with experts like Mr. Pasztor” and others working with WestExec, he said.
The American flag in the “government contracts” section of the pitch deck suggests these contacts could prove valuable later for other reasons as well. But Yedvab has denied these “initial conversations” are aimed at landing the U.S. as a client. Still, he noted to us, the importance of informing lawmakers of the existence of a “cost-effective” and “pragmatic” climate change option that Stardust thinks will soon be available.
The company earlier this year also hired law firm Holland & Knight to lobby policymakers. But the firm didn’t report its work for the startup until after we asked Stardust whether its representatives were making the rounds on Capitol Hill. The failure to initially disclose the $260,000 that Stardust has spent on lobbying so far this year was due to a clerical error, according to Holland & Knight spokesperson Olivia Hoch.
Stardust’s business plan and the resources and infrastructure aimed at making it happen were striking. But it’s normal for pitch decks and funding announcements to talk up the prospects of even the most outlandish business ideas. And Stardust’s latest funding round success shouldn’t be seen as a sure sign the company will succeed, either. Lowercarbon Capital, the venture fund co-founded by early Twitter and Uber investor Chris Sacca, led the funding round. The firm is highly respected for making big, contrarian climate bets that pay off, said Evan Caron, a fast-talking founder at the clean energy venture fund Montauk Capital. But many VCs still think it was a quixotic move to back a company that seemingly requires government contracts and international agreement to make any money.
“They’re kind of spraying and praying,” said Caron.
Lowercarbon defended its investment in Stardust, its first wager on a solar geoengineering company. “Stardust is the most rigorous scientific team in the world developing the capability to safely keep the planet from getting too hot for the most vulnerable humans and ecosystems,” said Clay Dumas, a founding partner at Lowercarbon who previously served as a special assistant in the Obama White House.
Indeed, despite the skepticism, some think it’s too early to dismiss Stardust. Aniket Shah, a managing director at the investment bank Jefferies, has authored several notes advising clients on movements in the geoengineering space. When he heard that Lowercarbon was involved with Stardust, he was stunned. “Wow, wow. That’s news, man,” he said. “That is an imprimatur that this thing’s going mainstream.”
Shah could see why investors were interested. “The energy transition is happening too slowly, and at some point, there’s going to be a response. And when that response happens from policymakers, largely, you want to own the company that can intervene,” he said.
Another Stardust investor, Maex Ament, a German fintech investor living in Madrid, said he had been won over by Yedvab’s sincerity and competence. In his mind, Stardust was “a project that I know the world needs.” But this wasn’t philanthropy, he was convinced there was a market. “I don’t have fuck you money. I’m not that rich that I just throw money left and right,” he said. Of course, he said, “there’s huge risk that maybe it will never be used. Sure, that’s startup risk.” But it came with a potentially huge payoff.
Stardust, he added, was the biggest investment he had ever made.
‘You don’t need to trust us’
In October, after months of polite deferrals, Yedvab agreed to speak to us for this story. He, his co-founder and a Washington-based media adviser flew all the way to London for a conversation in POLITICO’s Soho newsroom.
Yedvab, 54, looked every bit the accomplished scientist-turned-CEO. That is, winningly unpolished. His beard blended into loosely cropped salt-and-pepper hair. Alongside him was his co-founder and chief product officer Amyad Spector, a jumpy 42-year-old who was visibly protective of Yedvab. The pair, both physicists, had previously worked together in the Israeli government’s nuclear research division and had known each other for more than two decades. During the pandemic, they decided to do something about climate change. Their initial inquiries led them to despair. Until they came across the idea of sunlight reflection in a 2021 report from the U.S. National Academies of Science.
Yedvab said he had wanted to meet us face-to-face because “we have a lot to share.” He expressed a desire to be as open as he could, though he said the patent process constrained him. “We believe that eventually all the information should be out there,” he said.
But the key message, to which he returned repeatedly through a conversation lasting almost two hours, was his desire for governments to regulate the field.
“Stardust is not like any other company,” he said, because of the global implications of its technology and the “lack of a governance framework to make sure that we are doing the best we can.” His investors were betting that regulation would eventually come, and Stardust would be best positioned as a responsible actor to benefit from the government contracts they envisage.
“Our belief is that this field, realistically speaking, cannot move forward if you don’t have a very solid regulatory framework,” he said, speaking in good but imperfect English.
Though there was clearly another reason Yedvab was suddenly so keen to do an interview. Stardust was ready to announce the $60 million it had raised from 13 new investors. He was delighted, he said, not by the money, but what it meant for the project.
“We are, like, few years away from having the technology ready to a level that decisions can be taken” — meaning that deployment was still on track to potentially begin on the timeline laid out in the 2023 pitch deck. The money raised was enough to start “outdoor contained experiments” as soon as April, Yedvab said. These would test how their particles performed inside a plane flying at stratospheric heights, some 11 miles above the Earth’s surface.
Yedvab wouldn’t say what his particle was made from or where Stardust is seeking to patent it. But in an email, he said that the patenting process includes “a requirement to disclose all the details.” (In the U.S. and most other countries, a patent is only granted in exchange for the full public disclosure of the invention, generally within 18 months of the filing date.) All he would offer was that it was “built out of ingredients that are naturally occurring.” The key thing, he insisted, was the particle was “safe.” It would not damage the ozone layer and, when the particles fall back to Earth, they could be absorbed back into the biosphere, he said. Though it’s impossible to know this is true until the company releases its formula.
Yedvab said this round of testing would make Stardust’s technology ready to begin a staged process of full-scale, global deployment before the decade is over — as long as the company can secure a government client. To start, they would only try to stabilize global temperatures — in other words fly enough particles into the sky to counteract the steady rise in greenhouse gas levels — which would initially take a fleet of 100 planes.
If it were true that existing planes could be used, there would be no technical barrier for Stardust to begin operations, greatly reducing the timeline for deployment. But when this claim was relayed to MacMartin, the Cornell aerospace engineer, he was skeptical. Each flight would need to be able to carry many tons of particles to have any meaningful effect, MacMartin wrote in an email, adding, “There aren’t any existing aircraft that can get to 18km with a reasonable payload (and indeed very few that can get that high at all),” he said. In an email, Yedvab said: “According to our research, there is no showstopper, within existing aerial technologies.”
Even if the company’s particle proves to be safe and there are jets capable of the heights they claim, the question Stardust cannot answer is what happens if they get government contracts and go for it. There is only so much that modelling can tell us about how the Earth’s complex systems would react to humans toying with the thermostat. Yedvab’s answer was, essentially, that they would take it slow so they could easily dial down the dusting if things went wrong.
But, he admitted, there was no way to know every negative side effect without releasing the particles at scale. “The solution won’t be without consequences,” he said. He just wants them to be weighed against the potentially greater catastrophe of a superheated world. “Maybe there is a way out of the current corridor we are all walking in,” he said.
Each time he was probed on a safety issue, he would eventually offer the same response. The only way to discover whether solar geoengineering is a good idea is to do it; and only deep, careful regulation can make that happen safely. You don’t take medicines or get on airplanes unless they’ve been vetted by the feds, he argued.
We raised the point that the current state of the world is hardly ripe for a collaborative global effort to regulate solar geoengineering. To that, Yedvab said that the severity of the climate impacts meant it would be unethical to not make that effort. Dismissing something that could theoretically help “is not a privilege that I think that we as humanity have,” he said.
In his mind, the ideal scenario for deployment would be something like the international collaboration between nations and industry that saved the ozone layer from human pollutants through a 1987 U.N. deal — and he noted that this too was a pact forged amid geopolitical tensions. In the end, it shouldn’t be Stardust’s responsibility to decide whether its product is safe, he said. That’s the job of governments and international institutions.
“You don’t need to trust us. We are enablers of the technology, right?” Yedvab said. “But eventually someone else will need to test and validate and verify … You should build these layers that will bring trust.”
The last chance
It’s at least ironic that the plan for preventing Stardust from adding to human misery relies on the same beleaguered international system that has repeatedly come up short in its response to climate change.
But even as he moves into his delayed retirement, Pasztor’s work is finding new champions. Serious climate leaders in Europe, the U.S. and beyond — people who have taken his place at the center of the global diplomatic effort — are beginning to discuss building a regulatory response to Stardust and its ilk, according to conversations with several of those involved who were granted anonymity to speak about the confidential discussions.
Back by the shores of Lake Geneva in September, Pasztor took an outdoor seat at a casual bistro with white-and-black umbrellas and wooden tables beside the lake. It smelled like mowed grass. A steamship went past on its way to Lausanne.
The conversation about Stardust and Pasztor’s career had left him in a reflective mood. He’d just been to Colorado and spent a fortnight looking after his two grandchildren, who were seven and two. He has four in all, and he worries about them differently than he did about his own kids. “Our grandchildren are still helpless, they’re small,” he said. “They are just falling into this society.”
This wasn’t only the natural shift in perspective that comes from being a generation removed and a few decades older. It was between the births of Pasztor’s children and grandchildren that the world lost its chance to avert catastrophic global warming by ordinary means.
Thunder began bouncing off the mountains as we strolled back to his apartment. From the street, it was easy to see his balcony thanks to the Palestinian flag draped from the railing. Sitting back at his kitchen table in front of the now stale pastries, he talked about his years at the fore of the climate fight. The cycles of hope and despair. Now, he has found a new potential source of optimism in his work to ensure solar geoengineering is regulated properly. “I believe in it,” he said.
Despite the geopolitical challenges, Pasztor is still placing his faith in the collective wisdom and restraint of international cooperation. There’s simply no other option, he said, but to continue to compensate for the failures of the past with new attempts to get it right. “Until there is governance, these guys will do anything,” he said of Stardust.
The rain was now really coming down outside Pasztor’s window. Just behind the central square of the village was a park. Beneath the park, Pasztor said, was a bunker, one of roughly 360,000 public and private fallout shelters built by the Swiss during the Cold War in case of a nuclear strike. To this day, many are kept fully stocked with emergency supplies. It’s an enduring symbol of Swiss wealth and caution. It’s also an alpine hedge against human nature — a bet that someday soon all the safeguards will fail, and we will be faced with only bad choices.
Karl Mathiesen reported from Geneva, London and Brussels. Corbin Hiar reported from Washington. Laurens Cerulus contributed reporting from Brussels.
