Why and When Germline Editing is Permissible

"Case 4. Redefining Humanity

In 2023, the United Kingdom became the first country to approve a type of gene editing called germline editing, which allows scientists to change the DNA in sperm, eggs, or embryos to stop certain diseases from being passed down to children. By 2025, progress in science might make it possible to use this technology to fix more inherited health problems before a baby is even born.

Some companies, like Genomic Prediction, are going beyond just preventing illness. They’ve suggested that gene editing could be used to change traits like eye colour, height, and intelligence. They claim this could help advance human evolution. But many experts and critics are concerned. They warn that this could bring back ideas similar to eugenics—a way of thinking that promotes choosing certain kinds of people over others—and increase inequality in society.

Germline editing means changing the genetic code in ways that are passed down to future generations. New research projects like Genomics England, which studied 100,000 genomes from NHS patients, have helped scientists find over 600 gene changes that could be used to prevent diseases.

So far, gene editing has shown major success. For example, it has cured sickle cell anemia in 87% of patients in some trials. Scientists hope they can soon use similar methods to treat diseases like Huntington’s, cystic fibrosis, and muscular dystrophy before birth. Supporters compare this technology to vaccines andantibiotics—once controversial, but now widely accepted and trusted.

Some ethicists, like John Harris from the University of Manchester, believe it’s wrong to stop families fromusing gene editing to protect their children’s health. A 2023 Pew Research study found that 65% of parents support using gene editing to prevent disease.

However, others are more cautious. Right now, laws say gene editing can only be used for medical reasons, but companies like Genomic Prediction might find ways around that to offer changes for appearance or intelligence. In a 2024 study published in Nature, 72% of scientists said they were worried that this could lead to new forms of genetic discrimination or social pressure to create “better” children.

There are also concerns about how this technology affects people with disabilities. Organizations like the National Down Syndrome Society say that using gene editing to remove disabilities sends the message that people with disabilities are less valuable. In one survey, 84% of deaf people said they opposed gene editing to remove deafness, calling it a kind of cultural erasure. Many ethicists and activists argue that using gene editing this way could make people believe there is such a thing as a perfect person, which is a dangerous idea. As science continues to move forward, we need to ask ourselves whether we are just fixing genes—or changing what it means to be human.

1. Is it right to change genes that will affect future children who can’t give consent? What if the changes cause unexpected problems later in life?

2. CRISPR, the tool used for gene editing, still makes mistakes 15–40% of the time. Should scientists wait until it’s more accurate before using it on humans?

3. Should parents be allowed to fully control their child’s genes, or should there be rules to protect the child’s rights and future?"

POV: It is morally acceptable to edit genes for the purpose of curing diseases, although we should be cautious. This is because the child is a being who will be sentient in the future and will have its relational position with respect to human civilization as a morally important being, meaning that its welfare is important. While this may seem to lack consent or seem as a choice that disrespects autonomy, in fact gene editing does not involve violations of autonomy or disrespect of moral personhood, as it is just like a parent telling their kids what books to read or deciding what environment that the kid will be educated in. Autonomy is always imperfect.

When we examine the moral status of the child, two paths may assist in the analysis. From the property-based path, the child will be a conscious, sentient being who can suffer; thus, it is a human with moral personhood. From the relationship-based path, the genetically edited children will be and perhaps is already in relationships with respect to their parents and the human civilization, making their interests important. It may seem that relation-based ethics can more intuitively address the issue of whether the child has moral status – and indeed we will use relation-based ethics here – but both paths result in the same conclusion that genetically edited children are not objects to be used whenever but are beings worth moral consideration.

Even with that premise, however, it seems that gene editing is completely moral and perhaps even recommended in occasions where editing can help reduce the likelihood of diseases or disabilities. We tend to think that reducing severe diseases may increase the welfare of children, especially severe diseases like Down syndrome, Huntington's or cystic fibrosis. As we have argued, children are moral persons whose interests deserve consideration, their welfare should matter, and if genetic editing produces an overall increase in welfare, under welfare utilitarianism or hedonistic utilitarianism, genetic editing should be accepted as moral.

We might consider several objections:

The Autonomy Objection: Genetically editing children is morally impermissible because you cannot obtain the children's consent, which means that such an act violates the child's autonomy.

The key move of this objection is to offer a consideration of Kantian ethics, or at least a consideration of a practically pluralist framework. It invites autonomy into the scale and demands that we at least weigh the benefits in consequences against the loss of autonomy. Although I would happily concede towards a de facto pluralist framework where both utilitarian values and deontic values are seen as important, I believe that the act does not violate autonomy because autonomy is never perfect. To illustrate this response, we may consider an analogy.

Analogy to Parental Choices: Parents often decide where their children will grow and where they will be educated, sometimes even before their child was born. The parents may also choose to introduce some special hobbies to the children, which may allow him to grow fond of these hobbies. These such parental choices are not ultimately made autonomously by the children, for there is no possible way that the children could have consented to these choices. And yet, we tend to think that there is nothing morally wrong with parents deciding things for their children in such manners, especially when these choices are consequentially beneficial for the children.

The core reason that this exists is, as I aforementioned, that autonomy is never perfect. In our daily lives, we face choice architectures that are never neutral, and thus our choice is not at all completely rational, nor perfectly decided by ourselves. Instead, we are often affected by framing effects or other psychological imperfections that would alter or influence our reasoning in ways that we have no control over. Moreover, many things that we do or many thoughts that we have are connected to our our personalities, which in turn are affected by our growth environments, biological foundations, and early education, all of which are factors that we have no control over. Thus, one can easily see that even for the freest human individual, autonomy is never perfect in the ideal way of operating under self-legislating laws. Thus, all that matters is that it makes no difference between the level of autonomy that one enjoys after an act such as gene editing has been done, and the baseline level of autonomy that one would normally enjoy without that act. However, in the case of gene editing, adding or removing a gene does not really influence the level of autonomy enjoyed by that child in the future. The child had absolutely no autonomy over the genes that he would be born with even without the genetic editing, and unless one would believe that the randomness of the genetics is morally relevant (i.e. it matters whether the genes were chosen by nature or chance or by humans), genetic editing does not seem problematic, even if the framework is not purely consequential but rather pluralist.

Another objection comes from the risk of genetic editing:

Objection from the Risk of Genetic Editing: Genetic editing carries risks of harm, as it may fail in ways that demote the welfare of the edited. Moreover, implementing genetic editing means directly implementing a risk of harm as the necessary means to the desired end, which according to some (like those who believe in the Doctrine of Double Effect) may be completely unacceptable.

To this, the pluralist concession becomes particularly useful, for it allows us to import the doctrine of double effect, with intention redefined under Anscombe's terms, in order to understand why the risk of genetic editing only suggests caution, not impermissibility. The doctrine of double effect argues that only when harm is intended is it necessarily going to be immoral. In other words, if the harm is not an intended effect and is rather a foreseen but unintended side effect, it could be just to do the action. Looking at this in terms of Anscombe's definition of intention, where the mental motives and the psychological ideas of the agent doesn't matter, and where the action's composition and what its direct means requires is more important, we can clearly classify the risk brought by genetic editing, as long as the risk is relatively low, as a foreseen but unintended side effect, as the risk itself is not the direct means to the end of the genetic editing nor to the end of increasing the welfare of the child.

Moreover, this response can be strengthened by Thomas Rowe’s argument that the risk of harm itself is not necessarily a harm. Let’s consider a “Deadly Bingo” case, where Beth sets up a machine where one lottery ball will trigger a gun aimed at Adam. There is a real, objective chance that Adam will be shot. But Rowe argues that the risk itself still does not seem to harm Adam. First, the risk cannot harm Adam before the machine starts, because before activation there is no active risk. Second, the risk cannot clearly harm Adam while the balls are spinning, because the probability itself does not interfere with Adam’s interests. If a perfect predictor knew that the dangerous ball would not be drawn, then Adam would in fact escape unharmed. In that case, it seems odd to say that the risk itself damaged him, since nothing bad actually happens to his welfare. Third, the risk cannot harm Adam when the ball is drawn, because at that point the risk is over. Either Adam is shot, in which case the bullet harms him, or he is not shot, in which case he escapes unharmed. So the harm, if there is one, comes from the actual bad outcome, not from the probability of that outcome.

This matters because the objection from risk often assumes that, once genetic editing carries some risk, the child has already been harmed. But that seems too quick. If the editing is done for the sake of preventing a serious disease, and if the risk is reasonably low, then the mere existence of uncertainty does not show that the act is impermissible nor indicate harm. Instead, it shows that we need strict regulation, careful testing, and proportionality between the medical benefit and the possible side effects.

From the Objection from Risk, however, we can see that there are important qualifications to supporting genetic editing. In cases where genetic editing serves the purpose of changing biological features that are irrelevant to welfare, for instance, the color of a child's pupils, genetic editing may be less morally acceptable and perhaps even morally impermissible. This is because these features may not provide gains in welfare for the child and may even encourage eugenics – the idea that certain kinds of people are more preferred than others because of some characteristics or traits – potentially increasing discrimination and inequality, which we consider to be detrimental for societies. Moreover, when genetic traits are not edited for the child’s benefit, the child has no place in the end of this action, meaning that it has been effectively objectified, which constitutes a disrespect of moral personhood.

Q1. Is it right to change genes that will affect future children who can’t give consent? What if the changes cause unexpected problems later in life?

POV: Yes, even if the changes cause unexpected problems later in life. Arguments and objections have been discussed in the Main Stance.

Q2. CRISPR, the tool used for gene editing, still makes mistakes 15–40% of the time. Should scientists wait until it’s more accurate before using it on humans?

POV: No, using it is permissible, as per the Main Stance, but we should be very cautious until it is significantly lower, although there is no need for it to be perfectly flawless.

As per the Main Stance, risks are acceptable as long as the predicted likely outcome will mean increases in the child’s welfare. However, we indeed should be cautious about the risk, as 15-40% is very significant, and carefully consider the benefits and the potential harms. The government should only qualify certain hospitals which are technologically mature and are strongly supervised to operate. And provide clear and complete informed consent to those who undergo surgery. Some may use the Sorites Paradox to object against the caution:

Objection from the Sorites Paradox: It is difficult to locate one exact percentage of risk at which there is suddenly no need for caution. A decrease from 40% to 39% may not by itself transform the moral status of the act, just as removing one grain does not immediately turn a heap into a non-heap.

However, as Samuel Johnson has allegedly responded, the fact of twilight does not mean one cannot separate day from night. Though the borderline may be difficult, “wherever the line is, certain things are certainly well to one side or the other of it”. Clearly a 98% risk is radically different from a 1% risk; we should offer very much caution to a 98% risk while give considerably less caution to a 1% risk, although caution is certainly warranted if the negative outcomes from mistakes in genetic editing is so severe that even a 1% chance guarantees a mathematically expected outcome that is significant.

Q3. Should parents be allowed to fully control their child’s genes, or should there be rules to protect the child’s rights and future?

POV: Parents should have some control over their child’s genes – especially when they want to remove genes that could cause diseases, as per the Main Stance – but there also should be rules such that the child’s (future) moral personhood is also respected.

When genetic traits are not edited for the child’s benefit (e.g. pupil colors or hair colors are customized to fit the parents’ preference), the child has no place in the end of this action, meaning that it has been effectively objectified, which constitutes a disrespect of moral personhood. Moreover, as discussed in the Main Stance, such edits of genetic traits can promote the idea of eugenics – the idea that certain kinds of people are more preferred than others because of some characteristics or traits – potentially increasing discrimination and inequality, which we consider to be detrimental for societies.¹