Artificial Wombs Aren’t a Sci-Fi Horror Story

But that doesn’t make them any less meaningful of an advance.

An adult finger lifts up a baby's hand.
Getty Images Plus

Imagine a neonatal intensive care unit, maybe 10 years in the future. It’s a hushed, darkened room with rows of fluid-filled bags, each of them with a tiny baby inside. If it makes you shiver, if it makes you think of the battery farms in The Matrix or the hatcheries in Brave New World, you are not alone. This is a description of artificial wombs, and artificial wombs have baggage. But what this technology could actually do, and what we imagine it could mean, are two very different things.

The most common problem for premature babies is breathing, because lungs develop late in pregnancy. For fetuses in the womb, the placenta takes care of the oxygen­–carbon dioxide exchange. Clinicians keep premature babies alive with ventilators, which force air into lungs not ready to breathe, and incubators, which control their environment. This setup is an artificial womb, in a sense—it’s just nothing like a real womb. A proper artificial womb would bypass those immature lungs, reoxygenating the baby’s blood via an umbilical catheter, in an attempt to replicate the work of the placenta. And the baby would be held in fluid, floating just as they would in a biological womb.

It’s easy to let those baby-in-a-bag images take us to wild places, an inhumane future without mothers. A recent BBC story on a Dutch scientist working on artificial wombs put it like this: “It’s a very thin line between a dream come true and a horrific science fiction film.”

A more buttoned-up version of this sentiment was expressed in a recent letter to the editors of the American Journal of Obstetrics & Gynecology. It was a response to a study on the successful use of an artificial womb with premature lambs. The letter writer, a physician, wrote at length of his concerns, including his fear that “it might replace the role of a woman in natural pregnancy.” (The authors of the study later replied, “It is not possible to use this technology to supplant the role of women in natural pregnancy.”)

While the line between science fiction and reality is not, in this case, ambiguous, there are ethical and practical questions about the development of artificial wombs, as there are with every groundbreaking medical treatment. Among them: Who would have access to the treatment? How would nursing care change? What would be a child’s birthday: the day they were put in the bag or the day they were removed from it?

But the way that artificial wombs are spoken of suggests this technology—which has so far only been used with lambs—is on the cusp of changing everything about human connection and family and gender roles. It’s utopia or dystopia and nothing in between. These anxieties are preventing us from evaluating artificial wombs as they are actually being conceived. And they betray a lack of understanding of the urgent problem of premature birth.

In theory, it could work like this: In a modified C-section procedure, a premature baby between about 22 and 24 weeks would have three cannulas inserted into the three umbilical vessels (two arteries bringing spent blood out and one vein bringing freshly oxygenated blood back) to run their blood through a placentalike membrane that would provide oxygen and nutrition and remove carbon dioxide. Then they would be placed into a container filled with synthetic amniotic fluid. It would be a life support system modeled on the uterus, which is where premature babies are supposed to be. Basically: a baby in a large, fluid-filled bag with a thick tube where the umbilical cord should be.

It is no wonder that people find this disconcerting. It’s the uncanny valley of pregnancy, too close for comfort: the most intimate bodily process almost, but not quite, reproduced.

There have been notable recent successes on this path. In 2017, researchers at the Children’s Hospital of Philadelphia succeeded in keeping premature lambs alive and developing normally for up to 28 days in an artificial womb they call the Biobag. (Lambs are used because their respiratory systems are similar to those of human babies.) After removal from the bag, most were euthanized to study their organs, but at least one was allowed to reach adult sheephood, for continued study. And last summer, a group of Japanese and Australian researchers reported using their own device with extremely premature lambs, younger than those worked with in Philadelphia, closer to the 1-pound mark, the size of many 22- to 24-week babies.

Even this progress doesn’t mean we’ve recreated the womb. Scientists still don’t fully understand gestation: the labyrinthine interactions between the fetus’s and gestational parent’s bodies, the critical role of the placenta. If these new devices succeed, it would be incredible, but not because that labyrinth has been mapped and replicated. Artificial wombs wouldn’t be wombs so much as an improved version of the neonatal intensive care that already exists.

In NICUs today you’ll find a population of babies who are already partially gestated by technology: For the smallest babies, the first thing that usually happens in the delivery room is that they are intubated and connected to a ventilator. They are given nutrition intravenously at first, and then, if they do well, with a feeding tube that’s run down their esophagus. They live in an incubator, which keeps their bodies warm and their skin moist.

Hundreds of thousands of people are alive today because of neonatal intensive care. But it isn’t without a bodily cost, especially for those born before 25 weeks. These interventions can save their lives but also cause harm and discomfort: damaging their lungs and interfering with normal brain development. If artificial wombs work for humans, more babies will live, and with less pain and fewer health problems.

However, this technology, as currently conceived, would not allow humans to gestate entirely outside of the womb. Researchers say that before about 20 weeks, the blood vessels, heart, and skin of the fetus just aren’t developed enough to be supported in an artificial womb.

Matthew W. Kemp, a senior researcher on the study from Australia and Japan, explained it to me this way: “At the crudest level, if you can’t catheterize it and don’t have a fetal heart that’s strong enough to pump, that’s a hard limit. It’s really a hard stop, we think at 21 weeks, maybe a touch less, but certainly not lower than that.”

Kemp went on to explain in detail the limitations and highly specialized nature of this technology. “It’s not a replacement for a placenta or a uterus,” said Kemp. “The practical reality is that this is not a fun discretionary birthing option, like a water bath. Assuming for a moment that we are going to get this to work, it will be eye-wateringly expensive and require an extraordinarily skilled team of people.”

Pediatric bioethicist John Lantos compares the development of these devices to the slow change that has made it possible to treat younger premature babies with existing care. “I don’t think it’s going to change the nature of humanity, any more than saving 24-weekers changed the nature of humanity,” he said. “There have been significant shifts in what we thought of as viability. This could be another one. But it will be incremental, not cataclysmic.”

Today, in the United States, a fetus is considered viable if born somewhere between 22 and 24 weeks, though it depends on individual factors as much as gestational age. In that range, a premature baby has a chance.

The researchers who do this work say it would be impossible to use this technology with babies younger than about 21 weeks—and that pushing to earlier gestations is not their goal. That would mean these devices could shift the window of viability, but not in a dramatic way.

The concern over moving viability is mostly about abortion; the fact that we’ve defined the right to an abortion by a changeable line is a self-created problem (or rather, a problem foisted upon us by the anti-abortion movement). If we thought of abortion as a medical treatment instead of a subject for political and legislative debate, we would also be freer to evaluate the artificial womb, and other treatments for premature babies, on their own merits. Roe v. Wade established a cruel zero-sum game: Success in treating premature babies erodes access to abortion. What if we didn’t have to think of it that way? Canada, for instance, has no laws on abortion, because it’s a medical procedure, and there’s no need to legislate those like crimes. Otherwise, it’s hard to see why moving viability back, in and of itself, is a bad thing—if it means more treatments that help wanted pregnancies result in healthy babies.

This is a difficult climate in which to research pregnancy, and it has cost us. In the early 2000s, Helen Liu grew endometrial cells in her lab and then used that tissue as a scaffolding to grow mouse embryos. She hypothesized that she might be able to help people struggling with infertility as a result of implantation problems. It would be a bit like taking in vitro fertilization one step further and accomplishing implantation before transfer, then transferring the embryo and endometrium into the patient’s uterus. After quite a lot of press coverage, Liu actually halted her experiments. She told a reporter at the Atlantic that the pressure from the press and from anti-abortion and pro-choice advocates was just too much. Pro-choice advocates worry that if an embryo or fetus is “viable” at any time, legal abortion would disappear entirely. On the political right, the argument against artificial wombs centers more on religious or traditional ideas about pregnancy and motherhood.

Even outside of the abortion debate, the idea of what is “natural” seems to matter more when it comes to pregnancy and childbearing than it does in other medical contexts. The “natural” becomes a way to restrict or police women’s bodies and their choices. It can actively restrict progress in medical intervention in pregnancy, from abortion to IVF.

Is the artificial womb upsetting because we feel it is unnatural or because it comes too close to the natural, and complicates it? I think it’s both. Artificial wombs reveal the close-to-the-bone contradictions inherent in how we think about childbearing and motherhood.

When you see a 1-pound baby on a ventilator, in some sense, it feels wrong because it is wrong—not morally wrong, but wrong for their bodies. Premature babies’ bodies are meant to be in muffled, weightless fluid, not a loud, gravity-bound hospital room. They’re also not meant to be seen; they are meant to be sequestered in dark softness until they are ready to be among us. In this sense, the artificial womb is much more “natural” than what we have now. The salient comparison is to current neonatal life support, not to an actual body, a parent.

Natural vs. unnatural is an inadequate framework for the reality of trying to have a baby, a baby who lives. A fuller understanding of gestation—how we can or should intervene in it, and how we can’t—brings a more specific wonder and humility. If there are babies in bags, they will still just be babies. We will still be their parents. We will all be doing our best for them.

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.