This morning, a Dallas-based de-extinction company announced that twenty-six healthy chickens, ranging from a few days old to several months, have been hatched from a 3D-printed artificial eggshell. No biological shell was involved. The egg was built layer by layer in a printer, fitted with an engineered membrane that lets oxygen in and carbon dioxide out, fed calcium during incubation, and watched in real time through a clear window in the lid. The chicks came out healthy and look like chicks.
The company is Colossal Biosciences, best known for its work on bringing back the woolly mammoth, the dodo, and the Tasmanian tiger. The eggshell announcement is part of that broader project. The reason: you cannot bring back an extinct bird by laying its egg in a chicken. The egg is too big, the chemistry is wrong, and most extinct species don’t have a close living relative whose body could carry the embryo. If you want to resurrect the South Island giant moa (a 12-foot-tall flightless bird hunted to extinction by 1400), you need a way to grow a moa-sized egg outside of any bird’s body. That is what Colossal has now built the first version of.
What did they actually build?
The system has four parts:
- A 3D-printed lattice shell. Replaces the calcium-carbonate shell of a real egg. The geometry of the lattice is the result of months of work figuring out the right strength, the right porosity, and the right shape for the embryo to develop normally.
- An engineered membrane. Sits underneath the lattice. Lets oxygen in and carbon dioxide out at roughly the same rates a real eggshell does. This is a hard problem because the gas-exchange requirements change as the embryo grows.
- A calcium feed. A real embryo absorbs calcium from the inside of its eggshell as it builds its skeleton. The artificial shell doesn’t have any calcium to give up, so the team adds calcium to the inner contents during incubation, dosing it to match what the embryo needs.
- A clear observation window. The top of the egg is transparent. Scientists can take images of the embryo every few hours and watch it develop in real time. This is something no traditional egg permits, and it turns the eggshell into a developmental-biology research tool.
The process they used for these first hatches: crack open freshly laid chicken eggs, pour the entire contents (yolk, albumen, blastoderm with the early embryo, everything except the shell) into the artificial structure, seal it, drop it in an incubator, dose calcium on schedule, and wait 21 days. Twenty-six times, a chicken came out.
Why is this hard?
A bird embryo is one of the most delicate developmental systems in nature. The yolk has to stay oriented correctly so the embryo grows in the right position. Temperature has to be within a one-degree window. Humidity has to track a curve over 21 days. Gas exchange has to scale up as the chick grows from a dot of cells to a fully feathered bird. The eggshell handles all of this passively in a real egg, but it does it because hundreds of millions of years of evolution tuned its geometry. Replicating it in a printable, controllable system has been an open problem in developmental biology for decades.
Earlier attempts at “shell-less” chicken incubation, used in classrooms to let kids watch embryos develop, get the embryo to roughly day 17 of a 21-day cycle. They almost never produce a live chick. Colossal’s announcement is the first end-to-end success, the embryos make it all the way to hatching, and they hatch healthy.
What does this have to do with bringing back extinct species?
Three of Colossal’s flagship de-extinction projects involve birds or bird-adjacent species:
- The dodo. A pigeon relative from Mauritius, last seen in 1662. Colossal is working with the closest living relative (the Nicobar pigeon) to gene-edit toward a dodo-like genome.
- The South Island giant moa. The 12-foot-tall flightless bird of New Zealand, extinct by around 1400. No close living relative is anywhere near the right size. The egg alone is the size of a soccer ball. You cannot grow that embryo in a chicken or an ostrich.
- The elephant bird (similar problem to the moa).
An artificial eggshell solves the body-doesn’t-exist problem. If you can build any size of artificial egg, you can grow any size of bird embryo. The current system is chicken-sized. The team has stated in public that the design is variable, bigger lattices for bigger species, smaller for hummingbirds. The eventual moa shell would be the size of a basketball.
What else could this be used for?
De-extinction is the flashy use case but it is not the only one:
- Endangered species conservation. California condors, kakapos, and dozens of other endangered birds lay very few eggs per year. An artificial-shell system could grow embryos that current captive breeding cannot reliably support, and let researchers monitor development in ways they currently can’t.
- Drug and toxin screening. A clear-window egg lets researchers observe in real time how a chemical affects a developing embryo. This is a major use case in regulatory science (what’s safe for pregnant women, what causes birth defects).
- Developmental biology research. Watching embryo development in real time is a longstanding wish in the field. The artificial shell makes it routine.
- Poultry industry research. Vaccine development, breeding for disease resistance, and food-safety research all use chicken eggs as a baseline. A controllable, observable egg system speeds all of it up.
How does AI fit in?
AI plays three roles in the broader Colossal program (the artificial eggshell itself is mostly materials-science and physiology, but the surrounding pipeline is heavily AI-assisted):
- Genome reconstruction. Extinct species don’t have a complete, clean genome on file. AI is used to fill gaps in fragmented DNA recovered from museum specimens, comparing to the closest living relative’s genome to reconstruct what the original likely was.
- Edit prediction. Once you have a reconstructed target genome, you need to identify the specific edits to make in the living relative’s DNA. Machine learning models predict which edits will actually produce the desired traits and which will cause off-target damage.
- Image analysis during development. The real-time embryo imaging through the artificial-shell window produces an enormous amount of data. AI vision models track cell migration, organ formation, and developmental timing without a human watching the screen for 21 days.
The harder questions
De-extinction is one of the most polarizing topics in biotech. The arguments worth knowing:
- Is a gene-edited proxy really “the same species”? Colossal’s dodo will be a Nicobar pigeon with a set of edits that make it dodo-like, not a true dodo. Same with the mammoth (an edited Asian elephant). Critics call these “facsimiles.” Supporters say close enough is close enough for ecological purposes. The honest answer is somewhere in the middle and the science isn’t done.
- Is the ecosystem the species needs still there? The dodo lived in a Mauritius that had no rats, no monkeys, no pigs, no humans. None of those are gone. Even a real dodo would struggle to survive in modern Mauritius.
- Could the money do more good elsewhere? The standard critique: $235 million spent on de-extinction is $235 million not spent on preventing extinctions of species still alive. Colossal’s reply: the technologies developed for de-extinction (artificial wombs, gene editing, embryo imaging) directly benefit endangered species and human medicine.
- What happens if it works? If we can bring back the dodo and the moa, can we also bring back the woolly rhino, the giant ground sloth, the saber-toothed cat, or any of the dozens of species lost in the last 50,000 years? Where does the list stop, and who decides?
The Beginners in AI position
We are pro-technology. The artificial eggshell is genuinely beautiful work. Decades of biology questions get more tractable the moment a transparent, controllable egg becomes routine. Endangered species get a real lifeline. Drug screening for pregnant women becomes safer. And yes, we may eventually walk into a Mauritius nature reserve and see something that looks a lot like a dodo.
We are also pro-human first. The technology is at its best when it enhances life and judgment, not when it replaces them. “We can do this” is a different question from “we should do this.” The artificial eggshell is the same lesson the rest of the biotech frontier has been teaching us: the tools are now powerful enough that the deciding question is no longer the engineering. It’s the ethics, the ecology, and the values of the humans deciding what to build.
If a future version of this system saves the California condor from extinction, we will celebrate. If a different version is used to bring back a species we lost ten thousand years ago into an ecosystem that can’t hold it, we will write critically about that too. The technology doesn’t take a side. The humans do.
Frequently asked questions
Are these “really” chickens?
Yes. The contents of a freshly laid chicken egg (yolk, embryo, albumen) are 100 percent chicken. The artificial shell just replaces the calcium-carbonate housing. The chicks are genetic chickens with no edits, hatched in a non-biological container.
How is this different from the famous “shell-less chicken egg” classroom experiments?
Classroom shell-less experiments use a plastic cup with cling film and produce embryos that develop visibly but almost never hatch. Colossal’s system is engineered end-to-end: structural lattice, gas-exchange membrane, calcium dosing, controlled humidity and temperature. The chicks come out alive and healthy.
When will the first moa or dodo actually hatch?
Years away. The genome work and embryo work still have many steps to go. Colossal has publicly stated targets in the late 2020s for the dodo and the early 2030s for the moa. Those are targets, not promises.
Could this be used for poultry farming?
Probably not on a large scale anytime soon. A 3D-printed lattice is expensive compared to a hen laying an egg. The use cases are research, conservation, and de-extinction, not commercial chicken production.
Are there ethical concerns specific to this technique?
The specific technique is less ethically loaded than the downstream uses. Growing a chicken from a real chicken egg in a lab-built shell is not very different from incubating it under a hen. The harder questions arrive when this technique is paired with gene editing to grow species that don’t currently exist, or to bring back species that have been gone for centuries. Those are the de-extinction debates.
Where can I read the original Colossal announcement?
The announcement was carried by Associated Press and covered by National Geographic, Gizmodo, and dozens of outlets on May 19, 2026. Links in Sources below. Colossal has not yet published a peer-reviewed paper on this specific system. When they do, we’ll update this post.
Sources
- Colossal Biosciences, company website, public statements on the artificial eggshell project
- National Geographic: This chick hatched from an artificial egg (May 19, 2026)
- Gizmodo: De-Extinction Start-up Just Hatched Baby Chicks From a 3D-Printed Artificial Egg
- ABC News: A de-extinction company has hatched live chicks from an artificial eggshell
- Green Matters: Scientists Just Hatched Healthy Chicks From a Fully Artificial Egg