Alphabet is asking the EPA for 64 million Wolbachia mosquitoes in California and Florida

By David Ogilvy, Chief Marketing Officer at Mosticare Global | Published 2026-06-12

The technique is not new. What is new is the scale. Alphabet, the parent of Google, has asked the US Environmental Protection Agency for permission to release up to 64 million Wolbachia-infected male Aedes aegypti mosquitoes across California and Florida over the next two years — 16 million per state per year — as part of its long-running Debug project. The Federal Register notice landed on 6 May. The public comment period closed on 5 June. The EPA is now reading the room.

This is not the first time Alphabet has tried something like this. In 2017, its then-subsidiary Verily released twenty million sterilised male mosquitoes in Fresno, California, and watched the local Aedes aegypti population collapse in the release zones. Singapore — where the technique is called Project Wolbachia and has been running under the National Environment Agency since 2016 — has now published the most important numbers in the field: up to 98% suppression of Aedes aegypti at core study sites, and up to 88% reduction in dengue cases in areas that have hosted releases for a year or more.

So the EPA is not being asked to bless a moonshot. It is being asked whether a known, peer-reviewed, regulator-friendly mosquito control can scale to two of the largest US states.

What the technique actually is

Aedes aegypti is the mosquito behind dengue, Zika, yellow fever and chikungunya. Only the females bite, and they bite humans almost exclusively, which is what makes them such efficient disease vectors in cities. The Wolbachia method works because the females cannot tell the difference between a normal male and a male infected with the Wolbachia bacterium — they mate with both. When they mate with a Wolbachia-carrying male, the eggs do not hatch. The release males are sterilised in the lab (in some programmes by X-ray irradiation, in others by the Wolbachia infection itself, or both) so they cannot establish a population even if a female did manage to reproduce.

It is, in a word, mosquito contraception. The females waste their reproductive effort on sterile males; the next generation never gets born.

The technique is unrelated to gene drives, which modify the mosquito's own DNA to spread a trait through the population. Wolbachia is a naturally occurring bacterium already carried by roughly half of all insect species. The males being released are, biologically, normal mosquitoes with a passenger microbe. That is part of why regulators have been comfortable with it: the modified insects die out within a generation or two of releases stopping.

What the numbers in Singapore and Brazil actually show

The two big data points the field rests on come from tropical megacities where Aedes aegypti is endemic and the disease burden is real.

In Singapore, the NEA's Project Wolbachia has been releasing Wolbachia-carrying males across Housing and Development Board blocks since 2016, in partnership with Verily and the local firm Orinno Technology. By 2021, the agency reported up to 98% suppression of the local Aedes aegypti population at core study sites in Yishun, Tampines, Choa Chu Kang and Bukit Batok, with a corresponding reduction of up to 88% in dengue cases in those neighbourhoods. Associate Professor Ng Lee Ching, who runs the project for the NEA's Environmental Health Institute, called the result "very encouraging" and noted that it was the first time the technology had been shown to work in a "challenging tropical highly-urbanised and high-rise environment." The Duke-NUS dengue expert Professor Duane Gubler went further, telling the Straits Times that Singapore could "potentially be the first dengue endemic country in the world to effectively control this epidemic disease."

In Niterói, Brazil, the World Mosquito Program — a separate non-profit effort, also using Wolbachia but with a slightly different mechanism that establishes the bacterium in the local mosquito population rather than relying on continuous releases — published peer-reviewed results showing that Wolbachia deployment reduced dengue cases by 89% in the intervention area. (The RSI explainer in the brief cites this figure directly.)

These are not press-release numbers. They are published, peer-reviewed, regulator-reviewed. The honest caveat is that both are city-scale interventions under intensive operational management, and that scaling to two entire US states — even in selected neighbourhoods — is a different logistical question. The science is settled. The operations are not.

What the US proposal actually says

The Debug application, as described in the Federal Register notice and the press coverage that followed, is the more conservative of the two Wolbachia approaches. The males are sterilised in the lab. They are released, they mate, the eggs do not hatch, the population is locally suppressed, and the next generation carries no engineered trait. There is no ecological release of a self-spreading modification.

The numbers are large — 16 million per state per year, 64 million over two years combined — but they are not a sudden escalation. The 2017 Verily Fresno release was the proof-of-concept; subsequent Singapore and California work has refined the production line. The Debug application is, in operational terms, asking to do at state scale what has already been done at neighbourhood scale.

The entomologists quoted in the US coverage are broadly supportive but careful. Chris Grinter, an entomologist at the California Academy of Sciences, called the technique "really a genius" one. Nathan Burkett-Cadena, an ecologist at the University of Florida, drew the regulatory line precisely where it should be drawn: his concern was not the Aedes aegypti work, which targets a non-native species, but what would happen "if Google began to target native mosquito species." The mosquito that carries West Nile virus in the US, for instance, is a native Culex — and a Wolbachia programme against it would be a different conversation.

That distinction matters. Aedes aegypti is invasive in California and Florida. Suppressing an invasive species that is also a public-health threat is a fairly clean regulatory story. The moment a programme points at a native mosquito, the conversation gets longer, harder, and more honest.

The Swiss comparator and the regulatory cross-border

The Swiss Italian-language public radio piece in Martin's brief draws a useful line. Just south of the Alps, the SUPSI microbiology institute has been running its own sterile-male releases in the Ticino towns of Morcote, Ascona and Losone, with containment above 90% of the local tiger mosquito population. This year, SUPSI is extending the trial to Mendrisio. The Swiss technique uses radiation-sterilised larvae produced in an Italian lab and shipped across the border by post — a low-tech, low-cost version of the same logic.

The point of the comparison is not that Switzerland is further along. The point is that the Wolbachia / sterile-male approach is no longer one country's bet. It is being deployed, with regulatory permission, in equatorial megacities (Singapore, Niterói), temperate European towns (Ticino), and now being asked of the US. When the same technique is being tested across climates, regulatory regimes and at scales from a few thousand to tens of millions per year, the conversation stops being "does it work" and starts being "how do we run it well, and where do we draw the line on the species we point it at."

What to watch next

Three signals. First, the EPA's decision on the Debug application — and the conditions attached. A clean yes with monitoring requirements would be the most likely outcome, given the regulatory track record in Singapore and California. Second, what Alphabet actually does in year one. Releasing 16 million mosquitoes in a state is not the same as releasing 16 million mosquitoes in a neighbourhood, and the operational detail will tell us how seriously the company is taking its own pitch. Third, the Conversation of Native Species. The Burkett-Cadena caveat is the right one: Aedes aegypti is the clean case. Native Culex is the harder one. How regulators, scientists and companies handle the second conversation will tell us whether the Wolbachia moment is a one-decade technology or a one-century one.

The right reading of this story is not "tech company releases mosquitoes." It is "a mosquito control technique that has quietly been one of the most effective public-health tools of the decade is being asked to scale." Whether the regulator agrees, and how it agrees, will determine how many other cities get the same answer Singapore already has.

What we know

Sources cited

1. RSI (Swiss public radio) — "Lotta alle malattie negli USA, insetti sterili come in Ticino" (adapted by Sophie Iselin, RTS, 7 June 2026, 17:05). https://www.rsi.ch/audio/tematiche/scienza-e-tecnologia/Lotta-alle-malattie-negli-USA-insetti-sterili-come-in-Ticino--3803342.html
2. Smithsonian Magazine — "Google wants to release 32 million mosquitoes in California and Florida. Here's why." Sara Hashemi, 3 June 2026. https://www.smithsonianmag.com/smart-news/google-wants-to-release-32-million-mosquitoes-in-california-and-florida-heres-why-180988892/
3. The Guardian — "Google asks for permission to release mosquitoes in California and Florida," 1 June 2026. https://www.theguardian.com/technology/2026/jun/01/google-permission-release-mosquitoes-california-florida
4. National Environment Agency (Singapore) — "Project Wolbachia suppresses Aedes aegypti mosquito population and reduces dengue cases at release sites," 2021. https://www.nea.gov.sg/media/news/news/index/project-wolbachia-singapore-suppresses-aedes-aegypti-mosquito-population-and-reduces-dengue-cases-at-release-sites
5. Snopes fact-check — "Google, bacteria, and mosquitoes" (3 June 2026). https://www.snopes.com/news/2026/06/03/google-bacteria-mosquitoes/
6. Federal Register notice, 6 May 2026 (referenced by RSI and Smithsonian; govinfo.gov/content/pkg/FR-2026-05-06/pdf/2026-08808.pdf).
7. Martin content sweep, 12 June 2026 — item #2. intelligence/martin/2026-06-12-content-sweep.md