Solar Geoengineering: The Climate Fix Nobody Wants to Talk About

Solar geoengineering is the idea of deliberately cooling the planet by reflecting a small amount of sunlight back into space, and it has moved from a purely theoretical climate-policy debate to something private companies are actually testing in small amounts, right now, with limited government oversight. A 2026 US Government Accountability Office report lays out both the concept's real potential and how far ahead the technology has moved of the governance meant to manage it.

The two approaches actually being pursued

The concept splits into two main atmospheric methods. Stratospheric aerosol injection disperses reflective particles, typically sulfur dioxide, high in the atmosphere using balloons or aircraft, mimicking the temporary global cooling effect scientists have observed after large volcanic eruptions. Marine cloud brightening injects sea salt aerosol into low-altitude ocean clouds to make them more reflective, producing a more localized, regional cooling effect rather than a global one. Both remain largely unproven at meaningful scale — the GAO notes most of what's known comes from computer models, laboratory experiments, and a handful of small case studies, including one marine cloud brightening field trial conducted in Australia.

Private companies are already doing this in small amounts

Despite the technology's immature research status, at least two private companies are actively pursuing stratospheric approaches. One has reportedly been releasing sulfur dioxide since 2022. Another received $75 million in 2025 specifically to develop aircraft-based deployment technology. That's a meaningful gap: companies are operating in a space where the underlying climate science remains actively debated and where, per the GAO, formal government oversight is limited.

The GAO's central finding isn't that solar geoengineering is unsafe or safe — it's that the technology to attempt it has outpaced the governance frameworks meant to evaluate and regulate it, both domestically and internationally.

What could actually go wrong

The risks the GAO identifies aren't hypothetical hand-wringing — they're specific and significant. Stratospheric aerosol injection could affect air quality, damage the ozone layer, and alter precipitation patterns in ways that are difficult to predict regionally even if the global average effect is well modeled. Because the atmosphere doesn't respect borders, one country's or company's deployment could produce cooling or precipitation effects felt (and potentially opposed) in another country entirely — a geopolitical risk layered on top of the scientific uncertainty. Research also shows most surveyed Americans view solar geoengineering research negatively, and transparency concerns have already led to the cancellation of at least one planned experiment.

Why this connects to the rest of climate tech

Solar geoengineering is sometimes framed as a shortcut that makes emissions reduction less urgent, but the GAO's framing — and most serious researchers' — treats it as a potential emergency lever alongside, not instead of, reducing emissions and removing carbon directly. That puts it in an odd position relative to technologies like direct air capture, which addresses the cause (atmospheric CO2) rather than a symptom (how much sunlight reaches the surface), and carries a very different risk profile: DAC's biggest problems are cost and scale, while geoengineering's biggest problems are irreversibility and unintended regional consequences.

Where oversight actually stands

Federal oversight in the US remains limited, though several existing statutes could plausibly apply depending on how a deployment is structured, and some states have enacted their own bans. The GAO's report reads less as an endorsement or condemnation and more as a flag: policymakers don't yet have clear answers to what information they'd need to evaluate a deployment, or how domestic and international coordination could reduce the geopolitical risk of one actor's intervention affecting everyone else's climate. That governance gap, not the underlying atmospheric science, is the part of solar geoengineering most obviously unresolved heading into 2027.