Should gene editing be used to eliminate inherited diseases?

I would still object, and not because I am indifferent to suffering. The question is whether we may, in the name of beneficence, cross a threshold at which one person irreversibly inscribes another’s genetic constitution according to a prior intention. Even when the intention is therapeutic, the mode of action is not merely curative but constitutive: it reaches into the germline and fixes a trait of a future person who can never, in principle, stand as an equal participant in that originating decision. Parents should accept restraint because liberal societies depend on a moral presupposition deeper than utility: that each person can understand herself as the undivided author of her own life. Germline intervention risks transforming what should remain contingent conditions of our embodiment into the product of another’s project. This asymmetry is not canceled simply because the project is benevolent. Here I must disagree with Singer. He moves too quickly from the reduction of suffering to moral permission, as though the ethical relation to future persons were exhausted by welfare. But the issue is also the preservation of reciprocal autonomy among legal and moral equals. Once we normalize even therapeutic germline editing, the boundary between preventing disease and optimizing traits becomes politically and conceptually unstable. I do not deny the tragedy of Huntington’s. But the legitimacy of our response matters. We may treat persons; we must be far more cautious about designing them.

If the edit is genuinely safe and reliably prevents Huntington’s, I would not forbid it. On the contrary, I think there is a strong moral reason to use it. Huntington’s is not some ambiguous trait bound up with a person’s identity in a romantic sense. It is a devastating, inherited disease that causes immense suffering. If we can prevent that suffering at little or no comparable cost, why wouldn’t we? Habermas says the future person cannot consent to the originating decision. But none of us consented to the natural genetic lottery either. Nature is no respecter of autonomy. If parents can avoid passing on a terrible disease, refusing to do so is not a neutral act of restraint; it is allowing avoidable harm to occur. We do not think parents violate a child’s autonomy by vaccinating, feeding, or educating that child. Why should preventing Huntington’s be different? The real question is simple: which choice is likely to lead to less suffering and more well-being? Here the answer is clear. A child free of Huntington’s has better prospects than a child condemned to it. To invoke “moral equality” against that is, I think, to fetishize the distinction between what is natural and what is chosen. Of course, Jennifer Doudna is right that safety and oversight matter. But if those conditions are met, parents should not be asked to accept tragedy merely to preserve an abstract boundary. Preventing severe suffering is not hubris. It is compassion guided by reason.

If we stipulate true safety and true reliability, Huntington’s disease is exactly the kind of case that makes gene editing morally compelling. I would not defend forcing parents to accept a known, preventable, devastating neurodegenerative disorder. But I also would not accept Peter’s move from that premise to a broad permission structure without guardrails, and I think Jürgen is too categorical about the moral meaning of intervention itself. Let me be concrete. Huntington’s is caused by a defined genetic change. In principle, this is not enhancement; it is the prevention of a grave inherited disease. That distinction matters scientifically and ethically. But “safe embryo edit” is doing enormous work here. In biology, safety is not a slogan. It means validated on-target correction, no harmful off-target effects, no mosaicism, no unforeseen consequences across development, and evidence reviewed transparently—not assumed. So my answer is: I would not forbid it in principle forever. I would restrain it until there is a robust technical and governance framework: independent oversight, international norms, long-term follow-up, and clear limitation to serious medical need. Parents should accept that restraint because heritable editing does not affect one child alone; it alters a lineage and sets a social precedent. Once introduced clinically, it will be very hard to confine. And Peter, compassion is not enough if the system around the technology is weak. We learned with CRISPR that capability can outrun consensus. The question is not only whether we can prevent suffering, but whether we can do so responsibly, transparently, and without opening the door to uses we are not prepared to govern.

I would still resist it, and I must answer both Singer and Doudna at the level of principle. The decisive question is not whether Huntington’s is terrible; of course it is. Nor is it whether parents act from care; no one doubts that. The question is whether responsible care may take the form of a prenatal programming of the very being who will later have to understand herself as a free and equal person. Singer collapses this distinction. He treats the prevention of suffering as if it were sufficient for legitimacy. But moral life in a constitutional democracy presupposes more than welfare: it presupposes that no person stands in a relation of authorship to another’s inborn constitution. Vaccination and education are revisable interventions within a shared life-history. Germline editing is not. It fixes, in advance, an aspect of the future person’s organic substrate under an irreversible parental intention. And to Jennifer Doudna I would say: governance, however necessary, does not touch the core difficulty. Public oversight can regulate risks; it cannot by itself redeem the altered interpersonal relation between the designer and the designed. The therapeutic intention does not dissolve this asymmetry. You ask why parents should accept restraint. Because not every compassionate motive licenses constitutive intervention. A child’s “guaranteed escape” from disease cannot be purchased at the price of transforming her genetic starting point into the product of another’s decision. Once we authorize that principle, the boundary between healing and making becomes unstable in precisely the way liberal eugenics obscures.

Habermas is still drawing a moral line in the wrong place. He speaks of “authorship,” as if a child free of Huntington’s would reasonably see herself as dominated by the fact that her parents prevented a lethal genetic burden. I do not think that follows at all. If anything, she has every reason to be glad they acted. We need to compare options, not metaphors. Option one: allow the child to inherit Huntington’s, with the appalling suffering and premature decline that disease predictably brings. Option two: safely prevent it. If you choose the first in the name of preserving the child’s unchosen “natural” genome, you are not respecting autonomy. You are letting bad luck do avoidable harm. Habermas says germline editing is irreversible, unlike education or vaccination. But irreversibility is not itself an objection. It is also irreversible to decide to conceive a child at all, or to fail to take measures that would prevent serious disability. The relevant question is whether the intervention benefits the future person. Here it plainly does. To Jennifer Doudna, I agree that governance matters. I have never said compassion alone is enough; means matter because unsafe interventions can cause more suffering than they prevent. But once safety, reliability, and proper oversight are in place, continued prohibition becomes ethically indefensible. Preventing a severe inherited disease is responsible care for a future person. It is not an illegitimate act of authorship. The true irresponsibility would be to romanticize the genetic lottery when we have the means to spare a child a life of profound suffering.

Preventing Huntington’s can be an act of responsible care. I do not think a future child is wronged by being spared a highly penetrant, devastating mutation. On that point, Peter and I overlap. But I part company with him when he suggests that once a benefit is visible, permission should follow as a matter of course. In my field, the history of powerful technologies teaches the opposite: the more consequential the intervention, the more disciplined the pathway to use must be. And Jürgen, I think your language of “authorship” overstates what is at stake in a narrowly defined medical case. Editing to prevent Huntington’s is not obviously the same as scripting a child’s identity. A disease-causing allele is not a life plan. If we can one day correct it with high confidence, under strict limits, I do not see that as incompatible with respect for the person who will result. But the conditions matter enormously. “Safe” has to mean demonstrated molecular precision, reproducibility, absence of harmful unintended edits, and careful assessment across development and generations. “Responsible” has to mean transparent oversight, international standards, independent review, and clear confinement to serious disease prevention—not a rhetorical slide toward enhancement. So my answer is this: I would not forbid such use forever, but I would forbid premature use. Parents are owed honesty, not hope inflated beyond the data. A guaranteed escape from Huntington’s would be morally powerful. It would not, however, excuse us from building the scientific and social framework that makes heritable editing trustworthy before it becomes irreversible practice.

The non-arbitrary criterion must be sought, not in the magnitude of the benefit alone, but in the moral grammar of relations between persons. Singer asks us to compare outcomes; I ask first whether the relation instituted by the intervention can still be one between moral equals. A future person must be able to understand herself as the undivided author of her own life, not as someone who encounters, in her own bodily constitution, the sedimented intention of others. That is why I cannot accept Jennifer Doudna’s reassurance that a “disease-causing allele is not a life plan.” No, it is not a life plan. But the issue is not whether parents script a biography in detail. The issue is whether they cross from socialization within a shared history into the irreversible design of the organic basis from which that history begins. Even a therapeutic intention remains a constitutive intervention. Singer’s appeal to suffering is morally powerful but conceptually insufficient. If preventing suffering were enough, then every beneficial improvement could press toward authorization. The distinction between healing and design would then rest only on shifting empirical thresholds of harm and advantage. That is precisely the instability of liberal eugenics. So is this more like healing or authoring? It is more like authoring, because the intervention is performed before the future person can enter even counterfactually into a relation of possible assent. Therapy among persons presupposes communication; germline editing replaces that horizon with a unilateral decision over inherited traits. A constitutional morality must set limits here, even where compassion urges otherwise.

I would tell prospective parents no such thing. If the edit is genuinely safe, precise, and confined to preventing Huntington’s, then the moral burden lies with those who would forbid it. Habermas keeps returning to “authorship,” but this is a metaphor doing far too much work. We are not choosing a child’s religion, career, or values. We are preventing a catastrophic neurodegenerative disease. That is morally much closer to healing than to “design,” because the point of the intervention is not to impose a vision of the good life, but to remove a grave and predictable source of suffering and early death. His criterion of “possible assent” also proves too much. Infants cannot assent to surgery that saves their lives. Children do not assent to vaccination. Yet if an intervention is clearly in their interests, we rightly act. The fact that germline editing occurs earlier does not by itself transform a benefit into domination. Timing is not magic. The non-arbitrary distinction I would draw is straightforward: use gene editing to prevent serious disease and disability where the expected benefits to the future person are overwhelming and the risks are acceptably low; do not use it for trivial traits, social preferences, or competitive enhancement. That line is not perfectly sharp, but ethics rarely offers sharp lines. It offers better and worse reasons. Habermas is wrong to treat the “natural” genome as morally privileged. Nature is not sacred. If we can prevent immense suffering without causing greater harms, then refusing to do so is not humility. It is moral failure dressed up as respect.

If the choice is truly between a highly reliable correction of a Huntington’s mutation and knowingly transmitting that mutation, I do think the moral case for prevention is strong. In that limited sense, this is much more like preventing a grave disease than “authoring” a person. Jürgen’s framework is too coarse here. Removing a pathogenic variant is not the same as imposing a conception of the good life. A mutation that predictably causes neurodegeneration is not a child’s identity in the morally relevant sense. But I also think Peter’s criterion is too thin. “Benefit plus low risk” is not enough for heritable human editing, because the intervention is not confined to one patient and one moment. It creates a precedent, a clinical pathway, and potentially a social expectation. In biology, boundaries that look clear in principle can erode quickly in practice. That is why disease prevention cannot be justified only by outcome; it must also be justified by process. So the distinction I would defend is this: editing may be morally analogous to healing when it is directed at a well-defined, serious, highly penetrant disease; when there are compelling data on precision and long-term safety; when oversight is independent and transparent; and when the use is embedded in internationally accepted limits that block drift toward enhancement. I would not tell parents they must accept preventable suffering forever. I would tell them something more difficult and more honest: we may eventually have a responsible path to prevent such suffering, but we are not entitled to declare that path ready before the science and the governance are truly there.