A new US startup called Warhound has raised $200 million to genetically engineer working dogs for the military, police and first responders, and to equip them with protective gear1. The round values the company at $1.2 billion2. Its first stated goal is health: lowering the odds that dogs develop problems common to their breeds, such as hip dysplasia (a malformed hip joint) and certain cancers1. No edited dog has been born yet. The company expects the first puppies in early 20271.
What happened
Warhound launched with a funding round from a group of investors that includes Sequoia, 8VC and Corner Ventures, along with In-Q-Tel, a venture firm backed by the CIA and other US intelligence agencies2. The venture brings together Ben Lamm of Colossal Biosciences and Palmer Luckey of the defense company Anduril1. The two developed the idea after meeting a year ago2.
The plan has two parts. Scientists working for Colossal will change the dogs’ genes and then clone them1. Cloning means making a new animal that is a genetic copy of a donor, starting from one of its cells. Viagen, another company owned by Colossal, will handle the cloning using its existing technology2. Anduril will design gear and the software that runs it, such as goggles, vests and body armor, meant to protect the dogs while they work1.
Lamm told Wired that the company has already begun editing the genes of cell samples taken from its first group of donor dogs1. Those samples came from top-performing military, police and first responder dogs across the United States, mostly Belgian Malinois, along with some labradors and shepherds1. A separate account based on TechCrunch’s reporting names Belgian Malinois and German Shepherds as the breeds whose cell lines will be used2.
What the company wants to change in the dogs
Health comes first. The initial research will focus on cutting the risk of inherited breed problems, including hip dysplasia and certain cancers1. Lamm said many traditional working dogs fail training because of inherited issues like poor eyesight or hip dysplasia2. One report describes the goal as a new type of dog engineered for longevity, health and specific performance measures2.
Performance comes later. Lamm suggested to TechCrunch that further changes could follow, which might make the dogs faster and more agile or sharpen their sense of smell1. The dogs would then be trained for jobs working dogs already do, such as detecting bombs and helping in search-and-rescue operations1.
The company’s business case rests on a supply problem. According to Warhound, more than 80% of the US military’s working dogs are imported from Europe, and less than half of those imported dogs are fit for service when they arrive2. Those figures come from the company, not from an independent count.
The evidence so far, and where these reports come from
There are no results to judge. What exists is a funding round, a plan and a timeline. The company says gene editing has started in cells, not in living dogs1. The first batch is expected to be cloned and incubated soon, with births anticipated early next year, and the company aims to have the dogs in service within two years2.
Readers should also know how thin the reporting base is. The accounts Biohack reviewed are republished versions of two outlets’ stories. Several identical articles carry the line “Source: Engadget”13, and another is a rewrite credited to TechCrunch2. Those in turn quote Lamm’s interviews with Wired and TechCrunch1. Biohack did not have the original Engadget, Wired or TechCrunch pieces, or a statement from Warhound itself, so details here rest on those secondhand accounts.
What it does not show yet
- Whether the edits work as intended. No edited dog has been born, so there is no data on hip health, cancer rates, eyesight or smell.
- Whether the dogs live longer. Longevity is named as a goal2, but measuring it would take years after the first births.
- Which genes are being changed. The reports name the target problems, not the specific edits.
- How many dogs are planned, and at what cost. The exact pricing model is not yet public2.
Who is behind it: money, ownership and ties
The ownership is not quite what the public pairing suggests. Luckey is a partner in the business, but neither he nor Anduril holds a stake in the startup1. Colossal Biosciences does hold equity, and Lamm sits on Warhound’s board1. Colossal also licenses its technology to the new company2. That means Colossal stands to gain if Warhound succeeds, both as an owner and as a technology supplier.
Warhound’s chief executive is David Spirk, who most recently worked at Palantir1. Luckey founded Oculus VR, and Anduril partnered with OpenAI in 2024 to develop artificial intelligence for the Pentagon and has since won a US Army contract alongside Palantir1. Colossal is behind several “de-extinction” efforts, including one that claims to have revived the dire wolf and another that seeks to bring back the woolly mammoth1.
The target customers include the Department of Defense, police forces and first responders2. The reports do not describe plans for pet dogs.
Animal welfare and oversight
The company says it is working with the American Humane Society to certify its training program2. The group said it looks forward to ensuring the animals are treated humanely2. Warhound also plans to place retired dogs in reunification programs, giving them a path after their working lives2.
The reports do not say which regulators, if any, will review the gene edits or the cloning, or how the dogs’ health will be tracked and reported. Those questions remain open.
What to watch next
- The first births. Warhound expects its first engineered puppies in early 20271.
- Health data. Whether the company publishes how edited, cloned dogs compare with ordinary working dogs on hip problems, cancer and other inherited conditions.
- Service timeline. The company aims to have dogs working within two years2.
- Performance edits. Whether the company moves from health changes to the speed, agility and smell changes Lamm described1.
Biohack will follow this thread as the company shares more, and will report what is measured in the dogs once there are dogs to measure.
