Google Seeks Approval to Release 32 Million Mosquitoes in California

Google has asked federal regulators for permission to release up to 16 million lab-reared male mosquitoes in each of California and Florida every year for two years, a combined batch of roughly 32 million insects bred to knock down the mosquitoes that spread West Nile virus. The application sits with the United States Environmental Protection Agency (EPA, the federal agency that regulates pesticides), which is taking public comments through June 5 before it decides whether to grant an experimental use permit.

The pitch sounds novel, but Google has run a version of it before. Eight years ago the same program flooded two Fresno neighborhoods with sterile males and watched the local biting population collapse. The catch this time is the target: the company is aiming at a different mosquito, one tied to a disease that already circulates across California.

What Google Filed With the EPA

The filing, logged under experimental use permit number 92643-EUP-R, names Google LLC as the applicant. It covers live adult Culex quinquefasciatus male mosquitoes carrying a strain of Wolbachia pipientis, a bacterium found naturally in many insects. The releases would run across two states over two years, with a ceiling of 16 million treated males per state each year.

This is the latest move in Google’s Debug program, the mosquito-control effort Alphabet started more than a decade ago through its life-sciences arm, Verily. The earlier work focused on the insect behind dengue and Zika. The current request shifts the aim toward the species that carries West Nile virus and St. Louis encephalitis, two illnesses already established in California’s bird and mosquito populations.

Google says it would lean on artificial intelligence and robotics to breed, sort and release the insects at the scale the strategy needs. None of that has been approved yet, and regulators have not said where any release would happen. You can read the full notice in the EPA experimental use permit application for Wolbachia mosquitoes.

• 16 million treated male mosquitoes per state, per year
• 2 states in scope: California and Florida
• 2 years of releases under the proposed permit
• June 5 deadline for public comment to the EPA

How Wolbachia Turns Male Mosquitoes Into a Control Tool

The method does not poison anything. Wolbachia is a common bacterium that, when carried by a male mosquito, scrambles reproduction. When a treated male mates with a wild female that does not carry the same strain, the eggs she lays never hatch. Repeat that across a neighborhood for enough weeks, and the next generation thins out.

Crucially, only female mosquitoes bite people. The released insects are all male, so the plan would not add to the swarm that whines around a backyard at dusk. Here is the sequence Google’s system runs:

1. Breed Culex mosquitoes in a lab and infect the males with the Wolbachia strain.
2. Use automated sorting to pull out the biting females before any release.
3. Release the treated males into a target area during mosquito season.
4. Let those males mate with wild females, producing eggs that fail to hatch.
5. Track the falling population week over week and adjust the release rate.

The Fresno Trial That Set the Template

This is where the history matters. In 2017 Verily ran its first large American field test in Fresno County, releasing about one million sterile males a week for 20 weeks across two neighborhoods of roughly 300 acres each. It was the biggest release of Wolbachia-treated males the country had seen.

The 2017 Knockdown

Through the peak of that first season, the company reported a 68% drop in biting female Aedes aegypti inside the release zones compared with similar sites left alone. For a pilot, that was a strong signal that the approach could move the needle on a real population, not just a cage experiment.

The 2018 Scale-Up

The following year the team pushed harder and reported suppressing more than 90% of the biting female population at its field sites during the July-to-October peak. Independent academic work has since tried to verify how well these releases perform in the wild, including a peer-reviewed Bayesian evaluation of Wolbachia male mosquito releases against Aedes aegypti in Harris County, Texas.

The table below lines up that earlier work against what Google now wants to do.

Why Targeting Culex Raises the Degree of Difficulty

Switching from Aedes to Culex is not a small tweak. The Fresno success came against Aedes aegypti, a homebody mosquito that breeds in small containers close to where people live, which makes it easier to blanket with releases. Culex quinquefasciatus, the southern house mosquito behind most West Nile transmission, ranges more widely and breeds in messier water, from storm drains to neglected pools.

The disease stakes are higher, too. According to the Centers for Disease Control and Prevention (CDC, the federal public-health agency), West Nile virus is the most common mosquito-borne disease in the country, and it already circulates in California, where counties routinely report mosquitoes testing positive for West Nile virus. A positive sample turned up in Riverside County just before the proposal drew wider attention.

Google is not the only group hunting alternatives to spraying. Mosquito agencies have spent years testing tools beyond traditional pesticides, and researchers elsewhere are even developing a genetically modified fungus that lures and kills mosquitoes. The Wolbachia route has the longest track record at scale, but most of that record sits with Aedes, not the Culex this permit would target.

Public Comments, Local Skeptics, and What Approval Hinges On

The science is only half the hurdle. The other half is convincing the people who would host millions of extra insects, even harmless ones. In the Florida Keys, where a similar effort is in its second season, mosquito officials sound encouraged.

It’s a great concept, and we’re putting it to real use to see if it works.

That was Chad Huff, public information officer for the Florida Keys Mosquito Control District, speaking to KVUE. Not everyone is sold. Brent Nye, a Florida resident, told 10 Tampa Bay News he found the idea interesting but uneasy, saying he would rather the company test it somewhere else first because so many things could go wrong. That gap between expert confidence and backyard hesitation is exactly what the EPA comment window is built to surface before any insect leaves a lab.

Frequently Asked Questions

What Is Google’s Debug Mosquito Project?

Debug is a Google program, launched more than a decade ago through Alphabet’s life-sciences division Verily, that breeds and releases sterile male mosquitoes to shrink populations that spread disease. Its current EPA application is filed under Google LLC and targets the West Nile carrier Culex quinquefasciatus.

Will These Mosquitoes Bite People?

No. Only female mosquitoes bite, and the releases would be all male. Experts say adding treated males would not increase the number of biting insects, because the males’ job is to mate with wild females and produce eggs that never hatch.

Where in California Will the Mosquitoes Be Released?

Regulators have not announced any release sites. The EPA is still reviewing the application and has said it would decide on the experimental use permit before any locations are set in California or Florida.

How Can the Public Comment on the Proposal?

Comments go to the EPA under docket EPA-HQ-OPP-2025-3951 and must arrive on or before June 5. The agency posts the docket and instructions on the federal regulations.gov page for the mosquito permit application.

Is Wolbachia Safe for Humans and the Environment?

Wolbachia is a naturally occurring bacterium present in roughly half of all insect species, and it does not infect people or pets. The EPA has previously granted permits for Wolbachia mosquitoes; details on its earlier reviews sit on the EPA page on the Wolbachia mosquito permit.

Has This Method Worked Anywhere Else?

Yes. Beyond Google’s Fresno trial, the nonprofit World Mosquito Program reports large drops in dengue across multiple countries using a related Wolbachia approach; its figures appear on the World Mosquito Program impact dashboard. Results vary by mosquito species and local conditions.