Why Google is Eradicating the Mosquito

Kind: captions Language: en Every summer fleets of trucks cover American neighborhoods with chemical fog to protect against the deadliest animal on Earth, but it stopped working. Mosquitoes are evolving. They're resisting. Now diseases like West Nile virus are coming back. The era of chemical pest control is over and it's time for something drastic. Google wants to release [music] up to 32 million mosquitoes on purpose into neighborhoods across Florida and California. Not to spread disease, to end it. The mosquito earns its title as the world's deadliest animal. It's body count is impressive. >> [music] >> The diseases mosquitoes carry are estimated to cause as many as 830,000 deaths each year. That's more than the next top three killers combined. Humans at 430,000, snails, yes snails, at 200,000 and snakes at 138,000. In the United States, the most common mosquito-borne [music] disease is the West Nile virus. In 2025, the country recorded more than 2,100 cases and 172 deaths. There is no vaccine for it. There is no specific treatment. >> [music] >> There is only prevention and prevention means the fog truck. But those numbers don't tell the full story. Most infections never cause symptoms, never get tested, and never make it into official statistics. What we see is only the visible tip of a much larger outbreak moving through the population every summer. West Nile is carried by one branch of the mosquito [music] family, the Culex. The insect Google is actually going after is a different one and in global terms, it is far more dangerous. Its name is Aedes aegypti. This is the mosquito that carries dengue, Zika, chikungunya, and yellow fever. It is small, aggressive, bites during the day, and has adapted almost perfectly to living alongside humans. It can breed in something as small as a bottle cap of water inside your home. More than 2 billion people on the Earth now live at risk of viruses [music] it spreads and its range is expanding. It's pushing into California and entrenching across Florida. It is almost the ultimate enemy. It doesn't need a swamp or a marsh to procreate. A discarded tire, a clogged gutter, the saucer under a potted plant, a child's toy left out in the rain are enough to make it lay eggs. Anywhere a tablespoon of water can sit for a few days, Aedes aegypti can turn it into [music] a nursery. You can't drain its breeding sites because its breeding sites are everywhere you live. It's a evolved alongside human beings [music] and it treats our cities as its natural habitat. The diseases it carries are just as bad as West Nile, at least on a global scale. Dengue alone is estimated to cause between 100 and 400 million infections each year. Symptoms range from flu-like illness to severe hemorrhagic disease that kills over 10,000 people annually. Zika can pass from a pregnant woman to her child and damage [music] the developing brain. Chikungunya leaves survivors with joint pain that can linger for months. This is what a single insect delivers, one bite at a time. So, when Google says it wants to release 32 million mosquitoes into those two states, understand what's really being said. They are not introducing a new threat, they are targeting one that's already there. It already is spreading and it's already winning. They're doing it because the old way of fighting back has stopped working. No amount of fog trucks are going to save us. For most of the last century, the strategy against the mosquito was brute force. If the insect is the enemy, poison the air it flies through, spray the neighborhoods, fog the wetlands, coat the standing water. And for a while, this worked. Then the mosquitoes began to change. When you spray a population with a chemical, you don't really kill them all. You just kill the weak. The few that happen to carry a genetic quirk survive. Those survivors then breed, and their offspring inherit the quirk. Spray [music] again, and you create an even stronger breed. Within a handful of seasons, you are no longer poisoning a vulnerable population, you are evolving it. This is not a theory. Back in 2018, researchers in Florida tested the wild Aedes aegypti against permethrin, the workhorse of pesticide sprays. Every single population they examined was resistant. The most stubborn populations had become up to 61 times harder to kill than a mosquito with no resistance to the chemical at all. They then genotyped nearly 5,000 mosquitoes from more than 200 locations across Florida. And what they found was alarming. One of the key mutations behind insecticide resistance had already spread across almost the entire state. But the real shock came from the control group. A competing mosquito species breeding in the exact same containers showed almost no resistance at all. The chemical still worked, [music] just not on the mosquito that mattered the most. But there's another problem with the fog method. It doesn't stay where it's aimed. It drifts. It settles into gardens, vegetable plots, flowering plants that feed pollinators, and into beehives. A tool designed to kill an insect can't tell the difference between a bee or a mosquito. Instead, the bees that don't have resistance can die, while mosquitoes will need more and more of it to be culled into meaningful numbers. The strategy was failing twice over. It was becoming less effective against the target with every passing season. It was never precise enough to spare everything else. You were paying more, spraying more, and getting less while the collateral damage stayed the same. So, what do you do when the thing you're fighting evolves faster than you can invent new ways to kill it? Google's answer was to stop fighting biology with chemistry and start fighting biology with biology. To understand the solution, you have to understand something that is practically unique to the mosquito. Only the female mosquito bites, since she needs the protein found in blood to produce eggs. The male never bites at all. He feeds on nectar, lives a short life, and his only role is mate with the female. This is what the entire plan is built on. Flood a neighborhood with male mosquitoes that look and behave exactly like the locals, but they can't produce viable offspring. The population collapses from the inside, all without a single drop of poison. There was only ever one problem. To release males, you first have to separate them from females by the millions with almost no mistakes. Let too many biting females slip into the batch and you are not suppressing the problem, [music] but express shipping it to people's doorsteps. For most of the history of this idea, sorting mosquitoes by sex was painstaking manual work and that kept the whole approach small, slow, and local. This is where Google has an advantage. Verily, Google's life sciences division and a project called Debug attacked the problem the way Silicon Valley attacks everything, with automation. Mosquitoes are raised in tightly controlled facilities and then passed through computer vision systems similar to the technology that recognizes faces in photos. The machines identify each insect as male or female and then sorts them automatically. The process looks less like a laboratory and more like a factory line. The insects move in a single file past optical [music] sensors. In a fraction of a second, the system reads the physical signatures that distinguish a male from female. It spots the subtleties that a human technician would struggle to catch by eye over a long shift, and it makes the call. Males continue down the line toward release. Females are diverted out. It runs continuously at a pace no room full of people with microscopes could ever match because getting it wrong is not an option. So, how much better are the machines? One automated system can sort 150,000 mosquito pupae every hour while maintaining a claimed accuracy of 99.7%. [music] To match that output, two human workers would need more than an entire week. [music] In one Spanish mosquito control project, two workers manually sorted about 240,000 mosquitoes over [music] a full work week with an accuracy of roughly 99%. The scale difference is staggering. What once took days now takes hours. The mosquito has spent millions of years perfecting evasion. Now it has an opponent that doesn't get tired and improves every time it runs. Google's plan involves releasing millions of infected mosquitoes. >> [music] >> That word makes the idea sound far more terrifying than it actually is. There is no genetically modified bug, no lab-built DNA, no science fiction splice. The weapon is a bacterium called Wolbachia, and it is one of the most common microbes on the planet. It already lives naturally inside a huge fraction of the world's insects. Google's males simply carry a strain of it, and that is enough to break the species' ability to reproduce. >> [music] >> The mechanism behind the strategy is called cytoplasmic incompatibility. Males carrying a Wolbachia strain can mate normally, but if the female doesn't have a matching strain, [music] the embryo fails immediately. Fertilized eggs are produced, but nothing hatches. And because mosquitoes live such short lives, overwhelming a population with these males mean most females will lay eggs that go nowhere. Wolbachia isn't a single uniform organism. It is a single but highly diverse group of strains, and [music] different strains can affect closely related mosquito species in different ways. That means the system can be rotated. In one season, you can introduce one strain, [music] and the next you switch to another, continuing to cycle the incompatibility and suppression. If the idea of sterilizing a pest into collapse sounds far-fetched, it in fact has already saved North America from one of the most gruesome livestock plagues in its history. Decades ago, the United States [music] faced the screwworm, a flesh-eating fly whose larvae buried into living animals. So, scientists reared the flies, [music] sterilized the males, and then released them by the billions. Wild females mated with sterile males, produced no offspring, and the population was driven out of the country. The difference between that method and Google's is that Wolbachia adds a reach. Because the bacterium occurs naturally in so many insects already, it offers a way to achieve that same reproductive sabotage without irradiating the mosquitoes or engineering their genes. It's borrowing a tool that the natural world has already built and pointing it at a single species. And this particular version of the idea has a history with a fitting twist. The first time cytoplasmic incompatibility was used to wipe out a wild mosquito population, it wasn't even on the Aedes mosquitoes. Back in 1967, in what is now Myanmar, geneticist Hannes Laven used males from Culex pipiens fatigans, [music] a close relative of the West Nile mosquito. The result was a total collapse of the local population. And at the time, scientists [music] didn't even know the bacterium behind it was responsible. The same idea now being used against dengue mosquitoes was proven decades ago on [music] a close relative of the West Nile carrier. There's another aspect of the Wolbachia weapon that's important here. The bacteria make the mosquito less able to spread viruses. Some studies suggest that certain Wolbachia strains inhibit a virus from transferring to and from the mosquitoes. The research suggests newly infected mosquitoes are less likely to be able to spread disease. That last point gives another fail-safe. Simply introducing Wolbachia into a mosquito population can have another effect. Even if females carry it and avoid cytoplasmic incompatibility, they may become less able to transmit deadly viruses to humans. It sounds flawless on paper, but the real test is always going to be what happens when you open the box in an actual American town. Beginning in 2017 and expanding through 2018, the Debug team ran what was then the largest study of its kind in the United States. Sterile Wolbachia-carrying male Aedes aegypti were released across suburban neighborhoods in Fresno County, California. They were then compared against similar neighborhoods where nothing was released. By peak mosquito season, >> [music] >> the treated neighborhood saw a reduction of more than 95% in the number of biting female mosquitoes compared to the untreated areas. For every 20 biting females that should have been there, 19 were gone. In 2020, those results were published in the journal Nature Biotechnology, which meant the wider scientific community and regulators could pull the methods apart and check them line by line. The team even released its raw field data so the analysis could be replicated. This was no longer a press release, but evidence. The study had limits, however. The reason suppression didn't reach 100% is that biting females can wander from neighboring areas that were never treated. And that just means the technique worked so well inside the borders that its main remaining enemy was geography. And as such, if you treat a bigger area, the effect would only get stronger. Fresno wasn't the only proving ground. That same veraly built technology has been developed in Singapore's National Project Wolbachia. The city-state has been fighting back against dengue with repeated releases of sterile males. In far North Queensland, Australia, more than a million sterile males were released in partnership with the country's national science agency and local universities. The same underlying system has proved effective in different continents and different climates. There is a detail from those early Fresno trials worth holding on to because [music] it cuts against the instinctive fear of increasing the threat. Residents were shown demonstration boxes of the males being released and invited to let them land on bare skin. Nobody got bitten because male mosquitoes physically cannot bite. The insects being set loose by the millions are to a human being completely harmless. The only thing that fears them is a female mosquito looking for a mate. Google's plans for Florida and [music] California in 2026 are not about whether the plan works, but how to do it at the scale of an entire state. There's an idea in evolutionary biology called the Red Queen's hypothesis. It describes an arms race with no finish line where every advance by one side is answered by the other forever. For evolutionary biology, the concept applies to species adapting to their environment and being able to simply exist. The same concept has even been referred to in agriculture where it's [music] been used to describe how crops need to be modified to endure climate change. But that is exactly what chemical [music] pest control became. Every new pesticide purchased is good for a few years, then selected for the survivors who could shrug it off. Those survivors rebuild the population stronger than before. We were running as hard as we could inventing new poisons and the mosquito was running just as hard to survive. >> [music] >> The reason the Wolbachia approach is different is that it changes the playbook. A chemical attacks [music] the mosquito from the outside and the mosquito adapts. Wolbachia attacks it from the inside [music] through the one thing the mosquito can't stop doing if it wants to exist at all, reproducing. There is an elegant cruelty about it. And if it works at scale, it changes not just how we deal with mosquitoes, but how we think about managing the living world around us. Which is also exactly why the final step is the hardest [music] one. As of June 2026, Google doesn't yet have the green light. What it has is an application and a process. This is the moment it stops being about laboratories and field data and becomes about permission. The technique can be brilliant, [music] peer-reviewed, and proven across three continents and it still does not get to touch an American neighborhood until a regulator signs off. That gatekeeping is what this story is actually all about. A notice in the Federal Register laid out the plan. In the first year, up to 16 million mosquitoes would be released in Florida. In the second year, another 16 million in California. 32 million mosquitoes total. It's worth noting that this is a phased rollout across two seasons, [music] not a single mass release in both states at once. But that matter is is open to the public. Anyone can read it and submit comments through the [music] EPA's federal rule making portal under a specific docket number assigned to the project. This is how decisions like this are actually made, not in a boardroom, but in public view where regulators weigh the science and safety before anything is released. [music] And there's another side to the Red Queen's race. Whether this technology works in your country depends less on the science >> [music] >> than on your wealth, climate, and your government. Start with the fact that the world has already split into two camps. One approach releases both male and female mosquitoes carrying Wolbachia. The bacteria spread on its own and over time the protection becomes self-sustaining, gradually reducing mosquito populations. More than a dozen countries have taken this path. The other approach used by places like the United States, Singapore, and China releases only sterile males. It suppresses populations through brute reproductive disruption, >> [music] >> but the effect is not self-sustaining. That is the path Google is on and the one that is arguably more expensive because a suppression program is never finished. You need to keep manufacturing and releasing males season after season so you stop the mosquitoes from reproducing. In Singapore, that meant roughly two to five million male mosquitoes released every single week, delivered to treatment sites twice a week. In some cases, they were carried up to different floors of high-rise apartment blocks. It's basically a permanent subscription to mosquitoes, which the government has to pay for. [music] So, high-income places like Singapore or American states can afford the continuous male release method, but in mega cities where both the mosquito populations and the human populations are larger, the same approach becomes far harder and more expensive to sustain. [music] Then there's the climate trap. Sustained high temperatures can thin out the bacteria inside the insect, weaken the reproductive block, and blunt the very virus-fighting effect the whole strategy depends on. In other words, the tool tends to be the weakest in the hottest places. These are exactly the tropical heartlands where dengue does its worst work. Brazil alone accounted for half of the world's reports of dengue. All this means is that there's no single recipe that can simply be copied from a temperate suburb to an equatorial slum. And even when the money and the climate cooperate, there is the whole human bureaucracy. Health officials in many countries have never encountered anything like this. [music] The idea of deliberately releasing mosquitoes, some of which can bite, sits uneasily against a century of being told the only good mosquito is a dead one. Many governments have no regulatory pathway to even import [music] the insects. And the four to five-year rhythm of elections makes it hard to commit to an expensive program that might take years to show a visible result. A technology [music] can be proven and still stall in a committee room. So, the race does not actually end with Wolbachia. If approved, Google's million mosquitoes will be released into two of the wealthiest states on the earth with the laboratories, the money, and the regulators to make it work. Then, it's likely that there's going to be another project for more mosquitoes targeting the same [music] or even different states. Whether the same trick can eve