The Woolly Mammoth Returns as a Gene-Edited Elephant

Colossal Biosciences has pushed its first woolly mammoth-like calf into the early 2030s, after its gene-edit list grew past 150. CEO Ben Lamm said there is no hard date, not 2030 and not 2036.

The Dallas company, founded in 2021 with Harvard geneticist George Church, had aimed for a 2028 birth. The animal on the drawing board is still an Asian elephant with selected mammoth traits, not a clone from Ice Age tissue.

The 2028 Calf Moves Into the 2030s

Lamm’s new window lands after a year in which Colossal still talked publicly as if 2028 were in reach. In February 2026 he said the mammoth program was mostly done with editing and already more than two and a half years into elephant IVF and egg collection. By late 2026 he was telling interviewers the first birth belongs in the early 2030s.

The original plan was tighter than that. Colossal had pointed to a pregnancy in 2027 and a calf in 2028. Asian elephants carry a calf for 22 months, so even a perfect embryo tomorrow would not walk until the following year. The slip is not a pause in the lab. It is the calendar catching up with a genome that got harder the longer they stared at it.

THE SHIFT IN THE PUBLIC TIMELINE

1. 2021: Lamm and Church found Colossal with the woolly mammoth as the flagship.
2. 2027: Original target for a surrogate pregnancy.
3. 2028: Original target for a mammoth-like calf.
4. February 2026: Lamm says editing is mostly done and elephant IVF is well under way, while the company still cites 2028.
5. Late 2026: Lamm puts the first birth in the early 2030s, “not 2036, but not 2030 either.”

Each extra year matters because the next bottleneck is not software. It is an elephant pregnancy, a nucleus transfer, and a calf that has to live.

Twenty Wolf Edits Were the Easy Version

The company already ran a public dress rehearsal. In April 2025 it presented three gene-edited wolves as the first de-extinct animals, Romulus and Remus, born on 1 October 2024, and Khaleesi, born on 31 January 2025. Ancient DNA came from a 13,000-year-old tooth and a 72,000-year-old ear bone. The living genome was a gray wolf’s. Colossal rewrote 20 targeted sites across 14 genes and put no ancient DNA into those cells.

SOUND ON. You’re hearing the first howl of a dire wolf in over 10,000 years. Meet Romulus and Remus—the world’s first de-extinct animals, born on October 1, 2024.

The dire wolf has been extinct for over 10,000 years. These two wolves were brought back from extinction using… pic.twitter.com/wY4rdOVFRH

— Colossal Biosciences® (@colossal) April 7, 2025

The howl did the marketing. The biology did something quieter. A domestic dog carried the pregnancies, the edit count was small, and the result was a wolf the company said looked and sounded like a lost species. Paleogeneticists have a shorter description: a modified living relative. The mammoth program is that method with the volume turned up, and with a surrogate that is itself in trouble.

WOLF TRIAL VERSUS MAMMOTH PLAN

Once the wolf work is treated as a scale model, the mammoth delay reads less like a surprise. A 14-gene cosmetic pass is a different sport from rewriting more than 150 genes in an endangered giant.

A Shaggy Mouse Was the Test Run

Colossal did not start those 150-plus edits in elephants. Asian elephants are endangered, social, and slow to breed, so the company tried the coat and fat genes in laboratory mice first. Mouse pregnancy runs about 20 days. That is the whole point of the detour.

The woolly-mouse program, begun in September 2024, produced nearly 250 embryos across five rounds and implanted the better ones in about 12 females. 38 woolly mouse pups were born from October 2024, and Colossal says every one showed the long, wavy, gold-toned coat and the faster fat use the team was after. The public write-up counts 9 gene targets: seven for the coat, one for melanin, one for lipid metabolism. A Colossal preprint on the same work describes multiplex edits that recapture mammoth-like hair and fat traits in edited mice.

THE MOUSE PIPELINE IN BRIEF

• Gene targets: Nine edits covering coat texture, color, and fat use.
• Embryos: Nearly 250, built over five experimental rounds.
• Live births: 38 pups, all scored as showing the intended coat and fat traits.
• Clock: About 20 days in the mouse against 22 months in the elephant.

Lamm was unusually blunt about what that proves.

The woolly mouse project doesn’t bring us any closer to a mammoth, but it does validate the work we are doing on the path to a mammoth. It proves our end-to-end pipeline for de-extinction.

Ben Lamm, co-founder and CEO, Colossal Biosciences

The mice are healthy, he said, and they showed the team can stack several cold-related changes at once. They also showed the honest size of the remaining job. Hair and fat are the part of a mammoth you can photograph. Vasculature, skull shape, ear size, tail length, and how fat sits under the skin are still on the list, and that list is the reason 2028 moved.

An Endangered Elephant Would Have to Carry It

Colossal’s starting genome is the Asian elephant, which it says shares about 99.6% of its DNA with the woolly mammoth. The IUCN Red List lists the Asian elephant as Endangered, after an inferred decline of at least 50% over the last three generations. Collated 2018 estimates on that assessment put 48,323 to 51,680 animals in the wild.

That is the species whose cells are being rewritten, and the species whose womb, in the classic plan, would have to carry a hybrid fetus for 22 months. Zoo elephants already lose more pregnancies than animals in looser settings. A lab embryo with a mix of elephant and mammoth traits would not make that safer.

WHY THE SURROGATE IS THE CHOKE POINT

• One calf at a time: Elephants usually carry a single fetus through a 22-month pregnancy.
• Few animals to spare: The Asian elephant is Endangered, with tens of thousands left, not a livestock herd.
• Egg work already running: As of February 2026, Colossal had spent more than two and a half years on IVF and ovum retrieval.
• Scale problem: Lamm has said the company does not want a thousand elephant IVFs, and that artificial wombs have been on the plan from day one because it wants thousands of mammoths.

Church made the same arithmetic years ago. If cloning success is low, you need many fertile females for one live birth, and there is a limited number of fertile Asian elephants. Colossal’s process pages have also described implanting into African elephants, which are larger, a pivot that does not remove the welfare question. It only changes which endangered animal stands in the stall.

Chief science officer Beth Shapiro calls this a “both and” with ordinary conservation, not a substitute. The tools, she argues, can help living species too. The immediate cost still lands on elephants that did not volunteer, for a calf that will need a mother, a herd, and a diet no mammoth has taught since the species faded about 4,000 years ago.

Why the Edit List Keeps Getting Longer

Colossal first talked about changing about 60 to 65 genes in an elephant cell. That was the pitch that sat under the 2028 date. Lamm now says the job takes more than 150 genes, more than twice that first range. Traits that look simple from the outside, small ears, dense hair, a fat hump, sit on more DNA than the early maps showed.

The company has been mining that map from 59 woolly, Columbian, and steppe mammoth genomes, some about 3,500 years old and some more than 1.2 million years old. DNA in permafrost is broken. The work is comparison, not cloning. Machine-learning tools sort which differences look like they made a cold-adapted body, then CRISPR has to cut the living elephant genome at those spots, and someone still has to check that the rest of the genome did not move.

AI does not shrink the pregnancy. It only tries to stop the team from editing the wrong 150 sites. Each added cut is another chance to miss, to hit a neighbor gene, or to build a cell that will not make a viable embryo. That is ordinary genomics, and it is why a slipping date is what the work looks like when it is honest.

WHERE THE CLAIMS DIVERGE

• Colossal’s framing: A cold-resistant elephant with the core traits of a woolly mammoth counts as de-extinction if it can do the ecological job.
• The ethics critique: Arthur L. Caplan, a medical ethicist at NYU Grossman School of Medicine, writes that the methods create hybrids far from extinct species, and that a few dozen hairy elephants cannot rebalance a warming tundra on any serious clock.

In March 2026 Lamm said the team had already edited more than 25 genes into Asian elephant DNA, a long way from 150 and a long way from a uterus. Shapiro has talked about the first mammoth as if it were a couple of years off. The gene list grew anyway. Complex traits rarely answer to one switch, and the closer the ancient genomes are read, the more switches show up.

Investors Backed a $10.2 Billion De-Extinction Bet

The science is slow. The cap table is not. In January 2025 Colossal raised $200 million in Series C funding led by TWG Global, the vehicle of Mark Walter and Thomas Tull, at a $10.2 billion valuation. The company said that round took total capital raised to $435 million. Its Dallas write-up later added $120 million aimed at the dodo program and a conservation tie with the United Arab Emirates.

That money pays a 55,000-square-foot Dallas lab Colossal kept off maps, plus plans for another 30,000 square feet, a ViaGen cloning unit doing 200 to 300 cell transfers a day, and a roster that now runs to six species: mammoth, dire wolf, dodo, thylacine, bluebuck, and moa. The mammoth is the billboard. The platform underneath is multiplex editing, ancient-DNA assembly, and reproductive hardware that can be sold, spun out, or pointed at living endangered animals.

Lamm has said new species will move faster because the tools already exist. The dire wolf was the proof that a handful of edits can make a pressable animal. The mammoth is the proof that investors will wait, for a while, while the edit list doubles. Spin-out software and womb research can keep that story funded even if no six-ton calf arrives on the first advertised year.

The Calf Will Still Be an Edited Elephant

Colossal’s own finer print has been consistent when it is not selling a howl. The target is a cold-resistant elephant with mammoth hair, fat, and other Arctic traits. Company process notes have called the newborn a hybrid of mammoth traits on an Asian elephant genome. Intact mammoth cells do not exist to clone. What survives in Siberian ice is a puzzle, not a parent.

What has been done and is being planned is actually the creation of hybrids somewhat akin to but far from identical to extinct animal species.

Arthur L. Caplan, medical ethicist, NYU Grossman School of Medicine, in PLOS Biology

Caplan’s further point is practical. A handful of engineered elephants, born over years, cannot punch enough snow or eat enough dead grass to hold permafrost in a warming century. The first calf, if Lamm’s window holds, will be a social animal with no mammoth mother, no herd memory, and a genome its makers will still be arguing about. The live question after the wolf howl was where this method stops. The mammoth program has not answered it. It has only shown that the stop line, and the birth date, keep moving as the gene list gets longer.