Internationally recognised art and literary critic, writer and cultural producer. Published in ELLE, Glamour, Esquire and Cosmopolitan; member of the Union of Writers.
In Greek myth the aegis is the cloak or shield of Zeus, and later of Athena: a thing made by a craftsman, not grown, and lent to whoever holds power that day. It is a strange inheritance for a climate idea, but it fits. The proposal now circulating under various names, and which I will call the Aegis here, is a shield in space large enough to shade the Earth by about one per cent. It would sit near the first Lagrange point of the Sun and Earth, roughly 1.5 million kilometres sunward, where the two gravities balance and an object can hold station without much fuel.
None of that is new. What is new is that the idea has moved from the back of an envelope to a funded engineering survey, and that the survey says it is buildable with no new physics. That sentence should be read twice. It does not say affordable, wise or governable. It says that the thing we used to file under science fiction now has a budget line and a critical path.
The shield and the number
The arithmetic is simple and unforgiving. To cool the planet by about one degree Celsius, you must reflect around one per cent of incoming sunlight. Because the shield sits far from Earth, its shadow is not a sharp disc; it is a slight dimming spread across the whole globe. The Planetary Sunshade Institute, a US non-profit, published a baseline survey in October 2026 that compared six architectures. In one, about five thousand heliogyros, each roughly forty kilometres across, would together need some seventy million tonnes of material. That is around fifteen hundred times the total mass humanity has launched into space since Sputnik.
The report was funded by a grant of about £400,000 from ARIA, the UK’s Advanced Research and Invention Agency. That grant sits inside a £56.8 million programme called Exploring Climate Cooling, which funds twenty-two research teams and a handful of small outdoor experiments. ARIA is careful about what it is paying for. It says plainly that cutting emissions is the only sustainable answer, and that it is buying an evidence base so the world can decide, including deciding not to.
Two cost estimates bracket the argument. The institute refuses to give a figure, saying the uncertainty is too great. Its consortium member Chantal Cappelletti of Nottingham suggests tens of trillions of dollars. A separate NASA study, DimSun, proposes something different in kind: a controlled cloud of hollow dust particles at L1, reducing sunlight by 1.16 per cent, with a rough one-time cost of $134 billion including a 43 per cent margin. The dust would disperse within about a hundred days if the control spacecraft were switched off. That last property is the most interesting thing in the whole file.
Two ways to build a shadow
A sail and a dust cloud are not the same proposition. A sail is a structure: it must be manufactured, launched, assembled, station-kept and, eventually, dismantled. Its inertia is institutional as much as physical. A dust cloud is a process: it can be tuned up or down in days and switched off in a season. The 2026 survey notes that this reversibility is what most distinguishes space approaches from stratospheric aerosols, which stay in the atmosphere for a year or two and cannot be recalled.
Reversibility sounds like a safety feature. It is also a temptation. Anything that can be adjusted in days can be adjusted by whoever holds the controls, and the adjustment need not be uniform. A shield that dims the planet by one per cent is a global thermostat. Whether the northern hemisphere cools more than the southern, whether the monsoon arrives on time, whether high latitudes warm or cool, all of it depends on details of orbit, particle size and spectral choice. A shield can be aimed. That is a physics fact with a political consequence that no one has yet designed an institution for.
There is a second asymmetry. Cooling by reflection does not touch the cause. Carbon dioxide stays in the ocean and the air, and the moment the shading stops, the temperature returns to where the greenhouse gases say it should be, at a speed set by how long the shield has been up. Climate scientists call this the termination problem. A rapid stop after decades would be an experiment run in reverse on every ecosystem at once. The word “reversible”, applied to a planetary thermostat, means reversible for the operator, not for the frog.
What the institutions actually say
It would be easy to write this as a story about scientists versus activists. The record is more interesting than that.
In May 2026 the UN Environment Programme published a working paper synthesising the science. It did not call for deployment. It called for more research, and for governance that is inclusive and evidence-led. A coalition of academics and civil-society groups replied in July with a document titled “Precaution Means Non-Use”, arguing that the paper understated the justice problem and that funding research normalises the technology. The Convention on Biological Diversity has a de facto moratorium on outdoor geoengineering activity. The European Parliament’s research service briefed members in May 2026 and noted that the Parliament itself has called for an EU-wide moratorium on use and for a global framework built on a stance of non-deployment. The Council of the European Union published cautious conclusions in April.
So the field is not split between believers and deniers. It is split between people who want a governed research programme and people who believe that a research programme is the first stage of a deployment programme and should therefore be refused. Both groups are behaving rationally. That is why the argument will not be settled by more data. It is an argument about whether the existence of a tool changes the politics of using it.
And the tool is being built at a peculiar moment. The UK, through ARIA, is funding the research. The most detailed cost estimate comes from NASA’s innovative concepts programme. The most ambitious architecture comes from a private non-profit. No parliament has voted for a shield. No treaty names one. There is no agency that could license one, inspect one, or order it switched off.
What BRICS+ has to do with it
Here the story takes the turn that makes it more than an engineering essay. A solar shield is a public good in the technical sense: nobody can be excluded from its shade, and one country’s deployment changes the sky for everyone. Public goods of that kind are usually paid for by the strongest power, or by a coalition that agrees to be bound. Both options are in trouble.
The expanded BRICS group, which met in New Delhi in September 2026 under India’s chairship, is the obvious candidate for the second option. It contains the countries most exposed to heat, most dependent on monsoon rainfall, and most suspicious of rules written in Brussels and Washington. Its New Delhi Declaration talks about climate finance, resilience, critical minerals and a BRICS climate research platform. It does not mention a shield. I want to be exact about that, because a great deal of excited commentary has invented a BRICS space programme that does not exist. What exists is a bloc with the industrial capacity to make a shield, the motive to want one, and no mechanism to decide whether to have one.
That gap is the real subject. The countries that could build the thing are not the countries that would be most damaged by getting it wrong, and there is no forum where the second group can veto the first. This is the same problem the World AI Regulation Summit spent two days on in London this month: a technology arrives before its rulebook, and the rulebook is written by whoever got there first. At CC Forum in October I heard Jeffrey Sachs make the point about artificial intelligence in almost the same words, that the people building it warn it could be catastrophic and plan the initial public offering anyway. A shield is that sentence with a rocket attached.
The cultural objection
I have written about theatres, festivals and films for this magazine, and I want to say why this belongs here rather than in a science supplement.
A sky is not only a physical system. It is the oldest shared image we have. Every calendar, every harvest, every festival of light and every story about the gods begins with the assumption that the sun will do tomorrow what it did today. A shield does not break that promise, but it edits it. We would be the first generation to decide, deliberately and as policy, how much sun the next one gets. That is a cultural act before it is a technical one. It changes what we mean by nature, and it changes who we are in relation to it.
Artists have always known that a made sky is a political sky. Think of the painted ceilings of the baroque, where the light is carefully arranged above a throne. Think of the smoke over Victorian London, which Turner painted as beauty and which killed people. The question a shield raises is not whether humans can make a sky. We already have. It is who signs the design.
There is also a plainer objection, and I think it is the strongest one. A shield is a substitute for a decision we keep postponing. It would let the countries with the most historical emissions buy time, literally, while the countries with the least capacity to adapt live under a thermostat they did not set. The Maldives prosecutor at the AI summit said something in October that applies almost word for word: a small state has no market power, only the willingness to walk away. In a geoengineering regime there is nowhere to walk to. The sky is the one supplier with no competitor.
What would make it real
If the shield is ever built, three things will have happened first, and none of them is a launch.
The first is a measurement regime that everyone believes. Not a model: an observed, audited, independently verified record of what the shield is doing to regional rainfall and temperature. Without that, every drought for fifty years becomes a diplomatic incident. Reinsurance markets will not underwrite a planet with an unattributable climate, and insurers are usually the first people to price a risk honestly.
The second is a control regime with an off switch that does not belong to one state. The dust-cloud design makes this possible in principle and the sail design makes it expensive in practice. Either way, the question of who may reduce the shading, and on whose authority, is the whole ball game. A thermostat with no lock is not a public good. It is a hostage.
The third is a decision about the alternative. The honest case for research is that we may need a backup if decarbonisation stalls and tipping points arrive faster than adaptation can follow. The honest case against is that a backup changes the incentive to try. Both are true at once, which is why the argument is unresolvable at the level of principle and has to be settled at the level of institutions.
The 2026 survey puts construction twenty to forty years away, depending on how fast heavy-lift launch cadence grows and whether lunar material extraction becomes real. That is roughly the lifespan of a generation, or of a mortgage. It is long enough to build the governance first, if we choose to. It is also long enough to drift into deployment by a series of small, sensible decisions that nobody ever votes on as a whole.
The Greeks gave the aegis to a god because they understood that a shield is only as trustworthy as the hand that holds it. We are now designing one that will be held by a committee we have not appointed. That is the part of the project that needs the most work, and it is the part that no rocket can launch.
Maria Bregman writes on culture, climate and the institutions in between. This piece follows her reports from CC Forum London XIV and the first World AI Regulation Summit.
