Artificial womb technology exists today in a narrow but remarkable form: researchers have kept extremely premature lamb fetuses alive and growing in fluid-filled bags for up to four weeks, with organs developing normally and no signs of failure. No human baby has been placed in one yet, and the gap between a lamb experiment and a clinical device is enormous. But the technology is no longer science fiction. Several independent research programs are refining competing designs, regulatory agencies are weighing how to classify these devices, and ethicists are scrambling to address questions the law was never built to answer.
How an Artificial Womb Actually Works
The basic idea is to recreate, outside the body, the conditions a fetus experiences inside the uterus. That means solving three problems at once: gas exchange (getting oxygen in and carbon dioxide out without using lungs that aren’t ready to breathe), fluid environment (surrounding the fetus in a warm, sterile liquid that mimics amniotic fluid), and circulation (keeping blood flowing through the fetus and the gas-exchange device without the kind of mechanical pump that could damage fragile blood vessels).
The most well-known system, called EXTEND (EXTrauterine Environment for Neonatal Development), was developed at the Children’s Hospital of Philadelphia. It uses three main components: a pumpless circuit that connects to the umbilical vessels, a closed “Biobag” filled with a continuously exchanged fluid that acts as synthetic amniotic fluid, and an external membrane oxygenator that does the work a placenta would normally do. Because there is no mechanical pump, the fetus’s own heartbeat drives blood through the circuit, which keeps pressures gentle and avoids the vascular injuries that plague traditional life support for extremely premature newborns.1Nature Communications. An extra-uterine system to physiologically support the extreme premature lamb
A separate team in Australia and Japan developed the EVE (ex vivo uterine environment) system using similar principles but a different engineering approach. They tackled the problem of circuit resistance by using a smaller, lower-resistance oxygenator and administering vasodilators to the fetus to maintain organ blood flow. Early versions kept premature lambs alive for about 18 hours; later iterations with dual oxygenators extended that to 60 hours, and the most recent versions have achieved a full week of support in five of six premature lambs without infection.2PubMed Central. Development of an artificial placenta for support of premature infants: narrative review of the history, recent milestones, and future innovation
Why Extremely Premature Infants Are the First Target
Babies born at 22 to 24 weeks of gestation sit at the edge of what modern medicine can save. Their lungs are not ready to exchange gas, their blood vessels are fragile, and their brains are highly vulnerable to injury from the inflammation and pressure swings that come with mechanical ventilation. Survival rates at 22 weeks remain low even with the best neonatal intensive care, and survivors face high rates of chronic lung disease, brain injury, and developmental disability.
The promise of an artificial womb for these infants is that it could keep them in a fetal state rather than forcing them into a newborn state. Instead of intubating lungs that have barely formed and putting the infant on a ventilator, you would keep the lungs filled with fluid and let gas exchange happen through the umbilical circuit, just as it does in utero. The hope is that a few extra weeks of fetal-type development, protected from the harsh interventions of an intensive care unit, could dramatically improve outcomes.
Animal data supports this idea in encouraging ways. In lamb experiments, brains of animals supported on an artificial placenta showed significantly better white matter integrity than brains of animals placed on mechanical ventilation. Measures of brain tissue structure were roughly twice as high in the artificial-womb group compared to the ventilated group, suggesting that the gentler approach preserves developing neural pathways.3PubMed Central. Effects of an artificial placenta on brain development and injury in premature lambs
The Amniotic Fluid Problem
Getting the fluid environment right turns out to be one of the trickiest parts of the engineering challenge. Natural amniotic fluid is not just water with some salts dissolved in it. It contains proteins, growth factors, and other molecules that interact with fetal skin, lungs, and the gastrointestinal tract. A fetus swallows amniotic fluid continuously, which is critical for gut and lung development. If the synthetic version is wrong, those organs may not develop properly.
Recent animal work has shown that composition matters a great deal. When researchers tested different synthetic amniotic fluid formulations in a rodent model, standard saline solutions caused lung swelling, immune cell infiltration, and reduced airspace compared to controls. A more carefully formulated artificial amniotic fluid improved the production of surfactant proteins, which are essential for the lungs to function at birth. However, an even more concentrated version of that fluid actually suppressed those same proteins, demonstrating that getting the recipe right is a matter of precision, not just approximation.4PubMed Central. In Vivo Effect of a Synthetic Amniotic Fluid on Fetal Lung and Gastrointestinal Tract: A Pre-Clinical Rodent Model
There is also the question of the microbial environment. A natural uterus is not sterile in the way researchers once assumed. Some scientists have proposed that establishing beneficial microbial communities within an artificial womb, potentially using probiotic species, could help shape the developing immune system. But the research here is early and mostly theoretical. Questions about which strains to use, at what dose, and for how long remain unanswered.5PubMed. A critical review of the recent concept of artificial mechanical uterus design in relation to the maternal microbiome: An Update to past researches
Keeping a Fetal Circulatory System Running Outside the Body
A fetus circulates blood differently than a newborn. One of the key differences involves a vessel called the ductus arteriosus, which shunts blood away from the lungs (because the lungs are not being used for breathing) and toward the placenta. For an artificial womb to work, this fetal circulation pattern has to be maintained. If the ductus closes, which is what normally happens after birth, the whole system fails because the circuit depends on blood flowing through the umbilical vessels to the oxygenator and back.
Early pumpless artificial placenta experiments showed that this is possible but unstable. In one study, blood flow through the ductus arteriosus started at a healthy baseline, dipped when artificial support began, and recovered to about 84% of its original level. But toward the end of the experiments, the animals developed hemodynamic instability with declining blood flow in the aorta.6PubMed Central. Development of an artificial placenta I: pumpless arterio-venous extracorporeal life support in a neonatal sheep model This kind of gradual decline is one of the reasons these systems still have time limits, and one of the engineering frontiers that newer designs are working to push back.
The Gap Between Lambs and Human Babies
Premature lambs are a reasonable stand-in for human neonates in terms of size and lung development, which is why they are the standard animal model. But “reasonable” is not “identical.” A lamb fetus at the gestational equivalent of a 23-week human baby weighs considerably more, has somewhat different hemodynamics, and is physiologically adapted for a different species. Translating the circuit sizes, flow rates, and drug protocols to a human infant who might weigh 500 grams is not a simple scaling exercise.
Despite significant advances in these large-animal studies, no human trials have been conducted, and clinical translation remains speculative at this stage. A recent systematic review of preclinical evidence noted a critical gap between the promising animal results and the questions of viability, ethical governance, and regulatory readiness that need answers before any human infant could be placed in such a device.7PubMed Central. Artificial Womb Technology: A Systematic Review of Preclinical Evidence and Implications for Neonatal Viability and Intensive Care The technology could eventually influence viability thresholds at 22 to 24 weeks, but that influence is still theoretical.
What Regulators Are Wrestling With
One of the practical barriers to human trials is that nobody is entirely sure how to classify an artificial womb. In the United States, the FDA would almost certainly categorize it as a Class III device, the highest level of scrutiny, because it is life-sustaining. Class III devices require premarket approval based on sufficient evidence that the device is both safe and effective.8Baker Institute for Public Policy. Building a Framework for Artificial Womb Technology Clinical Trials
But the usual pathway for medical devices assumes you can test them in progressively larger trials with informed consent from the patient. The “patient” here is a fetus who cannot consent, and the stakes of failure are as high as they get. Ethicists have argued that clinical trials will be a necessary part of translation because regulators require them, but the conditions under which those trials should proceed remain deeply contested.9PubMed Central. Artificial womb technology and clinical translation: Innovative treatment or medical research? Should the first human candidates be infants who are already certain to die under conventional care? Should parents have to choose between a ventilator with known odds and an artificial womb with no clinical track record? These are not abstract hypotheticals. They are the questions that will shape trial protocols whenever the first applications are submitted.
Partial Versus Complete Ectogenesis
Almost all of the current research focuses on what is called partial ectogenesis: taking a fetus that has already been gestating inside a person and transferring it to an artificial environment for the remainder of development. This is essentially an advanced form of life support for extremely premature babies. Complete ectogenesis, growing a human from fertilization to term entirely outside a body, is a far more distant prospect and raises a different set of questions.
On the complete-ectogenesis front, the research is still at the cellular and tissue level. Scientists have recently developed lab-on-a-chip models that can recreate the earliest stage of human development, the moment when an embryo implants in the uterine lining. One group built a 3D system using human blastoids (lab-made embryo-like structures) co-cultured with bioengineered endometrial tissue, successfully recapitulating key implantation events and even modeling implantation failure.10PubMed. A 3D in vitro model for studying human implantation and implantation failure This work is valuable for understanding infertility and early pregnancy loss, but it is a long way from building a device that could sustain an embryo through nine months of development.
The vision of complete ectogenesis, using advances in artificial organs and ex vivo tissue culture, has been discussed in the literature for over a decade.11PubMed. The artificial womb But the gap between sustaining a late-term fetus for a few weeks (which current technology can do in animals) and replacing the entire gestational process (which no technology comes close to) is vast. Anyone claiming complete artificial gestation is imminent is overstating where the science is.
What Artificial Wombs Mean for Viability and Abortion Law
Legal frameworks around abortion in many jurisdictions rely on the concept of viability, the point at which a fetus could survive outside the womb. If artificial womb technology pushes that point earlier, even to 22 weeks or beyond, the legal landscape could shift in ways nobody has fully mapped out.
Legal scholars have already started flagging this. One analysis in the Michigan Journal of Gender and Law argues that artificial womb technology is a direct threat to the current understanding of viability, and that society needs to decide now how viability should be defined once such technology becomes mainstream. That paper ultimately argues viability should be retained as a legal standard but recommends legislative reform to place artificial wombs firmly within the private surrogacy sector, distancing them from being classified as medical apparatus that could be used to push the viability line earlier.12Michigan Journal of Gender & Law. How Viable is Viability? Artificial Womb Technology and the Threat to Abortion Access
The ethical literature on this is wide-ranging and unresolved. A scoping review identified concerns spanning from abstract questions about complete ectogenesis to immediate practical issues about what the authors call the “fetonate,” a fetal neonate being supported in an artificial womb. The review found both well-established and newly emerging ethical considerations, none with easy answers.13PubMed. Ethics Considerations Regarding Artificial Womb Technology for the Fetonate
The Moral Status Question
Perhaps the sharpest ethical debate centers on what an entity developing inside an artificial womb actually is. Is it a patient, like a premature infant in an incubator? Is it something new, a category that existing moral frameworks were not built for? Reasonable people disagree sharply.
Some ethicists argue that an artificial womb is just another piece of neonatal intensive care technology. Under that view, the developing entity deserves the same protections as any newborn being treated in a hospital. Others contend that artificial wombs are fundamentally different from incubators, that the entity inside them is neither a fetus (because it is not inside a person) nor a newborn (because it has not been born), and therefore represents a new moral category requiring new ethical reasoning.14PubMed Central. Considering the Developing Entity in an Artificial Womb as a Patient
This is not just a philosophical exercise. How you classify the entity determines who has decision-making authority, what level of medical intervention is required, and whether withdrawing support constitutes allowing a death, ending a treatment, or something else entirely. One reproductive rights advocate put the practical concern plainly: questions about responsibility are not just about doctors deciding what is and is not medically appropriate. They are also about the gestating person, or the person whose genetic material is involved, having a say in what happens to a fetus being developed outside anyone’s body.15Journal of Law and the Biosciences. ‘The law is very, very outdated and not keeping up with the technology’: novel forms of assisted gestation, legal challenges, and perspectives of reproductive rights advocates in England and Wales
Who Counts as a Parent
Existing parentage law in most countries was written with a basic assumption: someone gives birth, and that person is the legal mother. In partial ectogenesis, where a pregnant person delivers a fetus that is then transferred to an artificial womb, the definition might be straightforward. But complete ectogenesis, where no pregnancy occurs at all, breaks the model entirely. If an embryo is created from donor egg and donor sperm and gestated in a machine, the law in many jurisdictions has no mechanism for assigning legal parenthood.16Journal of Law and the Biosciences. ‘The law is very, very outdated and not keeping up with the technology’: novel forms of assisted gestation, legal challenges, and perspectives of reproductive rights advocates in England and Wales
This is reminiscent of problems that already exist in surrogacy law. In England and Wales, for instance, a surrogate’s consent to a parental order is not legally recognized until six weeks after birth. If an artificial womb replaces the surrogate entirely, there is no birth event in the traditional sense and no obvious person from whom to obtain consent. The legal scaffolding assumes a human body was involved at some point, and that assumption may not hold.
Religious traditions add their own layers of complexity. A Catholic moral analysis, for example, has argued that elective use of artificial womb technology is licit only when necessary to preserve the well-being of the developing entity or the mother, and that partial ectogenesis chosen for convenience rather than medical necessity deprives the conceptus of a right to be gestated in the mother’s womb.17PubMed Central. Artificial Womb Technology Cannot End the Abortion Debate for the Prolife Catholic Whether or not you share that framework, it illustrates that artificial wombs do not resolve existing moral debates so much as create new fronts within them.
Bonding, Psychology, and the Unknown Long-Term Effects
One concern that comes up in almost every public discussion of artificial wombs is bonding. The physical closeness of pregnancy, a fetus hearing the mother’s heartbeat, responding to her voice, sharing hormonal signals, is widely believed to play a role in parent-child attachment. If a fetus develops inside a plastic bag connected to machines in a hospital room, what happens to that bond?
The honest answer is that nobody knows. There is no collective understanding of the long-term psychological effects of artificial womb gestation on either parents or offspring. Researchers have flagged this as a major gap.18PubMed Central. Artificial womb: opportunities and challenges for public health The concept of “giving birth” itself changes in this context, and with it a whole web of cultural expectations, rituals, and identity. Parents of extremely premature infants already experience profound stress and disrupted bonding when their child is in a neonatal intensive care unit for months. An artificial womb might extend that separation, or it might reduce it by improving outcomes and shortening the overall ICU stay. Until the technology is used in humans and families are followed over time, the psychological dimensions remain a matter of speculation informed by adjacent experiences.
Equity and Access
If artificial womb technology works and becomes clinically available, who gets to use it? Cost-effectiveness and accessibility are among the major considerations that ethicists have identified. Neonatal intensive care is already staggeringly expensive. A device that combines membrane oxygenation, continuous fluid exchange, and round-the-clock monitoring would likely exceed the cost of even the most intensive current treatments, at least initially.
There is also the question of reproductive equity. If artificial womb technology is offered to cisgender heterosexual couples as a way to save extremely premature pregnancies, ethicists have argued there is a moral obligation grounded in social justice to ensure the same technology is available to sexual and gender minorities who wish to become parents. The history of assisted reproduction, surrogacy, and adoption shows a pattern of unequal access for LGBTQ+ individuals and couples. A review of that history concluded that, absent evidence of harm to children, equitable access to ectogenesis must be assured as a matter of reproductive justice.19PubMed. Equitable access to ectogenesis for sexual and gender minorities
This framing matters because it shapes how the technology is regulated from the start. If artificial wombs are classified strictly as emergency medical devices for premature infants, access will flow through hospital systems and insurance coverage. If they are eventually considered a reproductive technology, the regulatory pathway and the consumer base look entirely different. The classification decisions being made now, about whether these are medical devices or something else, will echo through who can use them and under what circumstances for decades.
Competing Research Programs and the Timeline
It is worth noting that this is not a single-lab endeavor. At least three major programs are developing artificial womb systems independently: the EXTEND team at the Children’s Hospital of Philadelphia, the EVE collaboration between Australian and Japanese institutions, and a group at the University of Michigan focused on artificial placenta designs. Each team has taken somewhat different engineering approaches, particularly around how to manage circuit resistance and infection risk.20PubMed Central. Development of an artificial placenta for support of premature infants: narrative review of the history, recent milestones, and future innovation
The EXTEND system has received the most public attention, partly because the 2017 Nature Communications paper included striking images of fetal lambs growing inside translucent Biobags.21Nature Communications. An extra-uterine system to physiologically support the extreme premature lamb The team has described maintaining fetal lambs, developmentally equivalent to extremely premature human infants, for up to four weeks with stable hemodynamics and normal growth.22Pediatric Research. The EXTEND system for extrauterine support of extremely premature neonates: opportunity and caution But the “opportunity and caution” framing in the title of a commentary on that work captures the scientific community’s mood. The results are genuinely impressive and genuinely insufficient for clinical use at the same time.
Realistic timelines for first-in-human trials are difficult to pin down. The FDA held an advisory committee meeting in 2023 to discuss the pathway for these devices, signaling that the agency considers the technology close enough to warrant serious regulatory attention. But “close enough to discuss” and “close enough to approve trials” are different things. Most researchers working in the field estimate that human trials are still years away, with some suggesting the better part of a decade before the regulatory, ethical, and engineering hurdles are cleared simultaneously.