Skip to Main Content

The startling announcement by He Jiankui almost one year ago that he had created the first genetically modified human beings unleashed a torrent of criticism. It also brought to the surface common misunderstandings — even among scientists and ethicists — that reproductive uses of this genome-modifying tool have therapeutic value, will treat people with genetic disorders, will save lives, and will eradicate disease. None of those are true.

The twin girls that He helped create are publicly known as Lulu and Nana. Their father is HIV-positive. The scientist said he used CRISPR-Cas9 genome editing technology to disable a gene called CCR5 to mimic a naturally occurring gene deletion that appears to confer immunity against HIV.

A major criticism from the scientific community, which has otherwise been generally supportive of advancing gene technologies, was that He did not use the technology to address a serious medical need. That criterion stems from a 2017 report by the U.S. National Academy of Sciences and the National Academy of Medicine recommending that, once the technology is ready and safe, genetic modification of embryos could be allowed when there is a “serious disease or condition” to be addressed and no “reasonable alternatives” exist.


This scientific criticism of He’s experiment was on target: Although HIV infection is a serious disease, there are proven ways to prevent transmission of the virus from an infected father to his offspring, and later in life to prevent or treat the infection.

That said, scientists criticizing He on the grounds that he did not treat a serious condition with no alternatives imply that other applications of CRISPR to modify eggs, sperm, or embryos could meet this standard. Indeed, many scientists hail reproductive applications of CRISPR (what I call rCRISPR) as potentially lifesaving and curative.


As I argued recently in the journal Bioethics, that claim for rCRISPR is mistaken and misleading. rCRISPR does not save lives or cure diseases that would otherwise exist. It merely aids in creating a person without a particular genetic disease who would otherwise not exist. The value here in creating healthy people — as opposed to healing people who would otherwise be sick — is nominal to none. If you are a genetically healthy person, you have a high chance of creating a healthy person right now. But nobody thinks you have an urgent moral reason to do so.

Let me be clear: I believe there are applications of CRISPR that do correctly purport to cure diseases in people who would otherwise have them. CRISPR-based gene therapies known as somatic applications, which are used in the bodies of people who have genetic diseases, have great curative potential. These aren’t the target of my critique. In fact, it would be much better if we invested the money spent on reproductive CRISPR applications on somatic ones instead.

Here’s why I say that using rCRISPR to modify eggs, sperm, and embryos does not cure disease.

Imagine an individual or couple at high risk for creating a child with a serious genetic disease. They have the following simplified range of options:

  1. Create a genetically related child in the time-honored fashion who will be at high risk for the genetic disease.
  2. Create a genetically related child using CRISPR who will be at very low risk for the genetic disease.
  3. Create no genetically related child.

The existence of option C undermines the claim that rCRISPR applications are lifesaving or curative. If the prospective parent did not have option B available to them, option A would not inevitably happen, meaning they would not inevitably have a sick child. Individuals have a choice in the matter of creating children at high risk of genetic disease: They can choose option C.

Here is a different way of seeing the point that rCRISPR is not morally urgent because it does not involve a child whose existence, or illness, is inevitable. Compare the following scenarios: A doctor creates a cure for a serious and debilitating disease and administers it to a seriously sick child, thus healing her. This doctor saved a life, a morally commendable action. Now imagine that the doctor first injected the disease into the child that he cured. Now his action in its totality could hardly be called curative, for he first created the harm. His seemingly heroic act of healing now has dubious moral value.

This is similar to what occurs with rCRISPR. A clinician first creates an embryo with genetic defects, and then “rescues” it using rCRISPR. Calling this a cure or a therapy is a nonstandard use of those concepts, one that warrants a distinction from standard cures that save lives and prevent otherwise probable harms.

The analogy shows that an otherwise good act or cure can have its moral value undermined if it is part of a greater act that caused the disease in the first place.

Saying that people at risk for passing on serious genetic disease to their offspring can simply not have genetic children at all may seem harsh. But there are several options available to such individuals for becoming parents. They could adopt a child, an often-overlooked opportunity to become a parent that also helps an existing child in need of a family. They could use the egg or sperm of a healthy donor to create a child. Both of these options are expensive, but not more so than rCRISPR.

Prospective parents could also create genetically related children using pre-implantation genetic diagnosis. In this process, an individual’s eggs or sperm are used in conjunction with a partner or donor’s gametes to create several embryos that are then genetically screened for the disease in question. Only healthy embryos are implanted for pregnancy. Pre-implantation genetic diagnosis is a safe technology that is being used today.

Reproductive applications of CRISPR simply cannot meet the criteria laid out by the U.S. national academies. They will not address a serious condition, because rCRISPR is used in the creation of a person who does not have to exist at all. Nonexistence is no condition, and thus not a serious one.

The real value of rCRISPR is not to the individual who is created but to the people who choose to use it to create a child. It helps individuals achieve their reproductive preferences by having genetically related children free of genetic diseases. But this preference, although perhaps deeply held by many, is not a medical need or an urgent one. And pre-implantation genetic diagnosis already enables people to meet this preference.

We should object to the investment of precious medical resources in bringing rCRISPR to market when there are serious and urgent medical needs that could be met elsewhere, including investing in somatic gene therapies that actually treat diseases.

Tina Rulli, Ph.D., is an assistant professor of philosophy at the University of California, Davis, and a member of the UC Davis Ethics Commons.

  • If you’d create a sick being just to cure it instead of creating it already cured, than you’re having some major problems and some mental issues, dear Tina. Nobody would ever do that when science allows to create a 100% healthy and illness resistant baby. It’s their future and the most important thing in their lives.

  • Bringing the future to the into the present. I’d use IVG and grow my baby in a lab away from the hubub and influences of the outside world (noise, chemical air polution, my eating habits, other interactions and so on) plus genetically edit her to witstand known illnesses and radiation (plus a few other things) so that when she will come out of that lab will be able to take and see the world from point zero. (Don’t talk about traits like how she’ll look like – nobody has and will never have the right to override your decission on that). […]

  • I agree mostly with argument for eggs and sperms. I am not an expert on development biology, but embryos can be up to 11th week of pregnancy (according to Google). If we are able to detect genetic disorders at the late embryo stage, wouldn’t rCRISPR be life saving? It seems early detection of genetic disorders and rCRISPR could be potentially highly valuable.

    • Hi Andrew. Thanks for reading. The technology I’m critiquing is only used in sperm, eggs, or embryos prior to implantation. You’re right, if we could intervene once an embryo was already implanted and a pregnancy had started, that would be a cure, i.e. potentially life-saving. And my argument would not apply. I think your question helps to highlight the crucial difference between pre-conceptive and post-conceptive measures. The former are simply not life-saving. To the latter, as to my knowledge, we are not doing this kind of intervention yet. Thanks again.

    • What a horribly arrogant article, to assume that those with inherited diseases could so easily adopt or go the IVF route. This article seems less against germline editing and more in favor of passive eugenics.

    • Hi Christy. I don’t see why rCRISPR will be any more accessible or affordable for most people than IVF or adoption. So if access is your concern, which I agree it should be, that is yet another argument against rCRISPR. My point is merely that those who could afford to avail themselves of rCRISPR could also afford IVF with a donor or PGD instead.

  • You make some interesting arguments. Some follow up questions/thoughts:
    1. Are you saying that when a couple has a child with a genetic disease or undergoes IVF to create embryos some of which have a genetic disease, somehow they and/or their IVF physician “caused harm” to that child/embryo by virtue of its creation?
    2. IVF w/ rCRISPR is potentially cheaper than gamete donation or adoption.
    3. By extension of your argument regarding the desire for genetically-related children, why do IVF at all for any indication? Are you saying that IVF in general, for infertility for example, is not ethically acceptable because patients can always adopt? Are you not discriminating against this group by denying their reproductive autonomy?
    4. Situations where rCRISPR is the only option to have a healthy child are admittedly rare. Is the fact that a disease or situation is rare justify not applying scientific or medical resources towards finding a therapy?
    5. If rCRISPR turns out to be as safe as PGD (admittedly a big if), I have trouble seeing why PGD is acceptable, whereas rCRISPR is not. Could it perhaps actually more ethical because it does not involve selecting against a particular type of embryo (potential future person)?
    6. Finally, prevention is almost always more cost-effective than treatment of existing disease and results in health care cost-savings down the road. Somatic gene therapy for example currently costs millions of dollars per year per patient vs ~ $50K for IVF/rCRISPR. There will always be people with genetic diseases of course because most people don’t know they are at risk (and many who do can’t afford or don’t choose IVF/PGD) and many are the result of de novo mutations. So, I am supportive of somatic gene therapy efforts as well.

    • Dr. Amato. A pleasure to discuss with you. To your questions.
      1. I am not making any claims about whether a harm is caused (although I think plausibly it is, and then it is cured. Just as when we amputate a leg to save a life). My only claim with creating an embryo and then curing it is that the moral value of the cure is undermined (in comparison with common cures) by the agent having caused the harm in the first place. That is, if the cure first requires causing the disease in the first place, then this cure is not morally urgent.
      2. My objection is to the investment of research dollars in bringing rCRISPR to market when there are existing alternatives. I don’t think investment cost is worth the money for redundant medical value. I’d be interested in how rCRISPR is cheaper than gamete donation or adoption. But it wouldn’t affect my main argument.
      3. If adoption were a more practically feasible alternative, I would make the very argument you suggest. But that’s an open empirical question. I should note however, that I have made the broader argument elsewhere that all prospective parents, including those pursuing IVF, but also the fertile, should adopt instead. My arguments are moral, not legal. I’m not interested in legally restricting the choices of the infertile. There’s a lot more to say here–for instance, whether reproductive autonomy requires the right to pursue reproduction any way one wants, or whether it is a negative right against noninterference. But again, my arguments are moral, not legal. So they entail no discrimination or restrictions against a group of people.
      4. A very rare group of people would benefit from rCRISPR and no other alternatives. But again, this is not treating a condition. It is enabling them to achieve a reproductive preference (to have genetically related children). So my argument is not just resting on the rarity of the condition. It is emphasizing that the “condition” is not a medical one, but a social preference. Plus it’s rare. Combined, those facts show that rCRISPR has little social value.
      5. Again, my objection is primarily to the investment of research dollars into a technology that is redundant in value to existing reproductive alternatives. If despite my objections it comes to market (which it will), then maybe it will be as ethical as PGD. But modifying genes to eradicate disease it itself a selection choice. If people use rCRISPR to modify a congenital disability, they are expressing a value judgment about the lives of certain people just as much as the use of PGD is.
      6. I mostly agree with this point. I just don’t think rCRISPR is preventative because it “treats” a person who need not otherwise exist. And, as you mentioned, it cannot eradicate de novo mutations.

      Thanks so much for engaging.

    • I want to edit what I said in 1 in the parenthetical. The leg case is not analogous. My mistake was made in haste. I don’t think creating people with lives worth living, with serious diseases, is a harm. But regardless, the question of harm is distinct from the point I was making–that the moral value of cure is undermined if the cure requires creating the harm in the first place.

  • If I understand you correctly, you value a currently living person more than that person’s genetically related offspring (if they have a genetically dominant defective gene). You assert that curing the disease in that person is ethical but curing it in that person’s gametes is waste of resources. That 1) they should just not have offspring – unless they use genetically superior donors, but 2) if they do happen to have kids, we should just keep fixing the mutation once the children are born? (Even if correcting the gene in one cell (a gamete) is easier than correcting the gene in a whole organism (an infant)?) Hmm. The first option (no kids) seems like old school eugenics and the second option (keep fixing the kids who are born) seems more wasteful than fixing the gametes. Reading between the lines of your essay, I would worry about the future you seem to espouse where there are “superdonors” with “superior genetics”. All of this is a slippery slope. Let’s try to choose a path that slides into a better future not one with less genetic diversity.

    • Hi Heather. Thanks for your comments. I just want to add that there are many options for having kids for person’s affected with serious genetic diseases. They can adopt. They can use a gamete donor. They can use PGD. So there are more than just the two options you lay out. And it is those very options that render investment in your option 2 unnecessary. Thanks for engaging.

    • I was thinking about what you said about old school eugenics and superior donors. One complication with your view is that using CRISPR on gametes to modify away disability/disease changes that persons gametes, so that their offspring will also inherit the change. rCRISPR is itself a eugenic technology. Somatic gene therapies in contrast usually do not affect the germline and are non-eugenic. So if eugenics is the concern, then one should not support rCRISPR.

    • Hi Anne, I’m a pro-choice feminist. I believe the government should not interference in my bodily rights. That’s entirely separate from saying that the government should invest millions/billions of dollars into creating reproductive technologies that let me achieve my reproductive preferences. I think there’s reasonable disagreement here on this latter point. But no inconsistency with feminism. Not giving people CRISPR (and instead spending that money on higher impact treatments) is not actively interfering with anyone’s bodily autonomy.

    • @Tina – fair point on public spending. I was more thinking about the recommendations from the national academies, which I take to concern legal prohibition, not just funding decisions.

  • How does it make sense to not cure a disease when it’s possible to? If a baby could be born without genetic diseases, overall humanity would have less diseases. People can still die from HIV and adopting children doesn’t stop it from spreading. If Crispr can cure the disease it won’t exist again in the bloodline therefore making the disease less prevalent.

    • Hi George. Good question. To eradicate genetic diseases, we would have to use rCRISPR in almost all of our reproduction (like in Gattaca). But proponents of rCRISPR are usually only advocating its limited use in certain high risk populations. They won’t wipe out all diseases this way. So I concede that you could use rCRISPR to significantly reduce disease if used in almost all reproduction. But many people objection to such widespread use. And such widespread use is practically infeasible. More details in my paper, linked in the article. Thanks for your comment.

    • I second Mr. Williams. Having fewer people be susceptible to disease is a good thing, even if all disease is not eradicated. Should we not rejoice in our great but incomplete progress in eradicating polio?

    • Hi Anne-I just want to emphasize that rCRISPR does not prevent disease in people who would otherwise have it. To use rCRISPR, you first create an embryo that would otherwise become a person with a disease and then “cure” it. Not using rCRISPR at all means no such person has to come into existence.

  • Anti-progressive attitude to the elimination of all genetic diseases in family lines. All limits on medical resources are due to the global unethical concentration of capital. Attacking and curtailing rCRISPR, working within this arbitrary system of limits imposed by irrational greed, isn’t the solution to the problem you present here. Radical redistribution under an enlightened command system is the solution.

    Lastly, by far the most prevalent and destructive diseases on the planet are those of the Schizopgrenia Syndrome. Curing psychosis, understanding the brain — this should be the priority of all medical science, if we prioritize by following the numbers, which wise leaders emphatically should.

    • Radical redistribution under command systems is a prevalent and destructive disease. Communism killed ~100M people in the 20th century. This is 50x more than all deaths from schizophrenia during that century (~2M).

Comments are closed.