In November 2018, a Chinese biophysicist named He Jiankui announced that twin girls had been born from embryos whose DNA he had edited using CRISPR-Cas9, making them the first known gene-edited humans. The experiment, which targeted a gene involved in HIV infection, was condemned almost universally by scientists, ethicists, and governments. It remains the single most consequential event in the public debate over human gene editing, and the fallout reshaped how the world thinks about what should and should not be done with this technology.
What He Jiankui Actually Did
He Jiankui recruited couples through a Beijing-based organization that helped Chinese people living with HIV. In each couple, the father was HIV-positive and the mother was not. Eight couples initially agreed to participate; one later withdrew. Of the remaining seven, five women had a total of thirteen embryos transferred for implantation, and two became pregnant. While the embryos were still in a lab dish during IVF, He injected them with a CRISPR construct designed to create a specific deletion in a gene called CCR5. The sperm had first been “washed” to separate it from semen, where HIV can be present. When each embryo was three to five days old, a few cells were removed and checked to see whether the edit had taken hold.1Journal of Law and the Biosciences. CRISPR’d babies: human germline genome editing in the ‘He Jiankui affair’
Twin girls, referred to by the pseudonyms Lulu and Nana, were born in late 2018. A third baby, from the second pregnancy, was reportedly born the following year. He announced the births at an international genomics summit in Hong Kong, apparently expecting acclaim. Instead, the reaction was swift and overwhelmingly negative. He was eventually sentenced to three years in a Chinese prison for illegal medical practice.
Why He Targeted CCR5
The gene He Jiankui tried to disable, CCR5, encodes a protein that sits on the surface of certain immune cells. HIV exploits this protein as a doorway: the virus latches onto CCR5 to gain entry into CD4+ T cells, the immune cells it ultimately destroys.2PubMed. The evolutionary history of the CCR5-Delta32 HIV-resistance mutation A naturally occurring mutation called CCR5-delta32, a 32-base-pair deletion in the gene’s coding region, prevents the receptor from being expressed on the cell surface. People who carry two copies of this deletion are largely resistant to the most common strains of HIV.3PubMed Central. The CCR5-Delta32 Genetic Polymorphism and HIV-1 Infection Susceptibility: a Meta-analysis
He’s stated goal was to recreate this natural deletion in the twins’ embryos. The logic sounded straightforward: if nature already produced an HIV-resistant version of the gene, why not engineer it? But this framing glossed over serious problems. The delta32 mutation occurs at meaningful frequencies only in people of European descent, and its effects in other genetic backgrounds are not well understood. More fundamentally, there were already safe, reliable methods to prevent HIV transmission from an HIV-positive father to his children during IVF. The sperm-washing step He himself used is one of them. The medical justification for the experiment was, at best, thin.
CCR5 Does More Than Let HIV In
One of the less obvious problems with disabling CCR5 is that the gene does not exist solely to serve as an HIV entry point. CCR5 plays active roles throughout the body, particularly in the brain. Research now shows that CCR5 is involved in regulating neuroinflammation, synaptic plasticity, and cognitive function. Both genetic and drug-based inhibition of CCR5 have been linked to reduced activation of immune cells in the brain, better preservation of the blood-brain barrier, and improved cognitive performance in various experimental and clinical models.4PubMed. CCR5 as a key modulator in neurocognitive disorders
That might sound like disabling CCR5 could be a cognitive boost, and some early coverage speculated exactly that. But the picture is more complicated. CCR5 is part of the chemokine signaling system that directs immune cells to sites of infection and inflammation throughout the body. Knocking it out entirely, from before birth and in every cell, creates unpredictable trade-offs. The gene may offer protection against certain infections beyond HIV or play roles in immune responses that scientists have not yet fully characterized. Modifying the gene’s population-level effects is, as one analysis put it, “extremely difficult to model,” and the medical risks of altering the gene pool in this way could be substantial.5PubMed Central. Risks and benefits of human germline genome editing: An ethical analysis
The Technical Risks of Editing Human Embryos
Even setting the ethics aside, the technology was not ready for what He attempted. CRISPR-Cas9 works by using a short guide RNA to direct an enzyme to a specific stretch of DNA, where it cuts both strands. The cell then repairs the break, either by stitching the loose ends together (which often introduces small insertions or deletions) or by using a template to make a precise repair.6PubMed Central. Mechanism and Applications of CRISPR/Cas-9-Mediated Genome Editing This system is powerful but imprecise in ways that matter enormously when applied to embryos.
The first major problem is off-target editing. The guide RNA can bind to DNA sequences that are similar but not identical to the intended target, causing the Cas9 enzyme to cut in the wrong place. Research in zebrafish has shown that when embryos are edited with CRISPR, roughly a quarter of the offspring can carry off-target mutations, and about 9% carry large structural changes at those unintended sites. These are not subtle single-letter typos; they include insertions and deletions of 50 or more base pairs that can disrupt genes far from the intended edit.7PubMed Central. CRISPR-Cas9 induces large structural variants at on-target and off-target sites in vivo that segregate across generations Newer editing tools like base editors also generate off-target structural variations in mouse embryos, suggesting the problem is not unique to the original Cas9 approach.8PubMed Central. Adenine base editors induce off-target structure variations in mouse embryos and primary human T cells
The second problem is mosaicism. When CRISPR is injected into an early embryo, the enzyme does not edit every cell at the same moment. Some cells get edited, others do not, and still others get edited differently. As the embryo divides, this creates a patchwork of genetically distinct cells within the same organism. In mouse embryos, continued CRISPR activity across successive cell divisions expands this patchwork further, with retroelement insertions adding to the allelic diversity.9PubMed Central. Retroelement Insertion in a CRISPR/Cas9 Editing Site in the Early Embryo Intensifies Genetic Mosaicism A mosaic person might have the intended edit in some tissues but not others, or might carry different unintended edits in different organs. The consequences are impossible to predict ahead of time, and extremely difficult to detect after the fact.
Available evidence suggests that Lulu and Nana are indeed mosaic. One of the twins reportedly has the intended CCR5 deletion on only one copy of the gene, not both, meaning she likely has no HIV resistance at all. The other twin may carry different edits on each copy, neither of which perfectly replicates the natural delta32 mutation. The experiment did not achieve its own stated objective.
The Ethical Failures
The scientific community’s condemnation was not just about the technology being premature. It was about a cascade of ethical violations in how the experiment was conducted. Analyses of He’s study identified problems at every level: questionable scientific value given the availability of existing HIV-prevention methods, an unreasonable risk-to-benefit ratio, an illegitimate ethics review process, invalid informed consent from the participating couples, and what amounted to regulatory misconduct.10PubMed Central. Experiments that led to the first gene-edited babies: the ethical failings and the urgent need for better governance
The consent process drew particular scrutiny. An ethical analysis of the consent materials used in the study found serious deficiencies in how risks were communicated to participants and whether the couples could meaningfully understand what they were agreeing to.11PubMed Central. The Consent Form in the Chinese CRISPR Study: In Search of Ethical Gene Editing The couples were HIV-affected families who may have been especially motivated by the promise of protecting their children, making the power dynamics of the researcher-participant relationship fraught. The institutional ethics board that supposedly approved the experiment has been widely questioned, with reports suggesting its review was superficial or possibly fabricated.
Perhaps the most fundamental ethical problem, though, is that germline editing affects people who cannot consent: the children born from the experiment, and every generation that follows them. Unlike somatic gene therapy, which modifies cells in a living patient and stays with that patient, germline edits are incorporated into eggs and sperm. They are inherited. Any mistake, any off-target effect, any unforeseen consequence becomes part of the human gene pool permanently.
Germline Editing Versus Somatic Therapy
The distinction between germline and somatic gene editing is the central dividing line in both the science and the ethics of this field. Somatic editing changes DNA in specific cells of a living person. If you edit blood stem cells to treat sickle cell disease, the change stays in those blood cells and is not passed to any future children. Germline editing changes DNA in an embryo, egg, or sperm cell, meaning the edit will be present in every cell of the resulting person and can be inherited by their descendants.
Public attitudes tend to reflect this distinction. Surveys conducted before and after the CRISPR era consistently show high approval for gene editing used to cure serious diseases, even when that editing involves the germline of embryos. But genetic enhancement, using the technology to improve traits beyond normal health, is rejected by citizens worldwide.12Frontiers in Genome Editing. What do people think about genetic engineering? A systematic review of questionnaire surveys before and after the introduction of CRISPR The line between therapy and enhancement is fuzzy in practice, but the public draws it firmly in principle. Among students surveyed in Germany, roughly half remained in favor of germline editing even after being presented with detailed scenarios, though concerns about enhancement, interference with nature, and loss of human uniqueness were commonly raised.13PubMed Central. Students’ attitudes towards somatic genome editing versus genome editing of the germline using an example of familial leukemia
Some ethicists argue that the therapy-versus-enhancement framing is itself a distraction. The kinds of edits currently achievable by gene editing are nowhere near sophisticated enough to manipulate complex traits like intelligence or athleticism, which involve hundreds or thousands of genes interacting with each other and the environment. The more pressing unresolved questions involve safety, consent, equity of access, and governance, not the specter of designer babies.14PubMed Central. Moving Beyond ‘Therapy’ and ‘Enhancement’ in the Ethics of Gene Editing
CRISPR Therapies Already Approved for Patients
While the CRISPR-baby scandal dominated headlines, somatic CRISPR therapies moved quietly through clinical trials and into approved use. The most prominent example is Casgevy (exagamglogene autotemcel), approved by the FDA in late 2023 for treating severe sickle cell disease. The therapy works by editing a patient’s own blood stem cells outside the body, then infusing them back. The edit boosts production of fetal hemoglobin, a form of hemoglobin that substitutes for the defective sickle hemoglobin and prevents the painful and dangerous vaso-occlusive crises that define the disease.15PubMed Central. Revolutionary breakthrough: FDA approves CASGEVY, the first CRISPR/Cas9 gene therapy for sickle cell disease
The clinical trial results were striking. Of 30 patients who had been followed long enough to evaluate, 29 were free from vaso-occlusive crises for at least twelve consecutive months, and all 30 were free from hospitalizations for those crises over the same period.16PubMed. Exagamglogene Autotemcel for Severe Sickle Cell Disease For patients who previously spent days or weeks in the hospital each year managing pain episodes, those numbers represent a transformation. The therapy is somatic: the edits remain in the patient’s blood-forming cells and do not enter the germline.
Casgevy illustrates why many scientists consider somatic editing the appropriate near-term application of CRISPR technology. The patient is a consenting individual. The edits are confined to specific cells. The risks are borne by the same person who stands to benefit. And the editing takes place in a dish, where cells can be quality-checked before being returned to the body, something impossible with embryo editing.
The Global Regulatory Response
He Jiankui’s experiment happened in a regulatory gray zone, and the aftermath prompted governments to take stock of where they stood. A comprehensive survey of 106 countries found that 96 had some form of policy document relevant to editing human embryos, gametes, or their precursor cells. Of those, 75 countries prohibit using genetically modified embryos to start a pregnancy. No country explicitly permits heritable human genome editing.17PubMed. Human Germ Line and Heritable Genome Editing: The Global Policy Landscape
But the patchwork nature of these policies is itself a problem. Many countries have no specific legislation addressing gene editing at all, relying instead on general regulations around assisted reproduction or medical research. And even among the 75 countries that prohibit heritable editing, five provide exceptions to their prohibitions. The lack of binding international governance is a recurring concern among scientists and governance professionals, who have called for globally recognized boundaries and coordinating frameworks that do not yet exist.18PubMed Central. Human Heritable Genome Editing and its Governance: Views of Scientists and Governance Professionals Without those frameworks, the worry is that another researcher in another country with lax oversight could repeat what He did.
Newer and Potentially Safer Editing Tools
Part of the technical objection to He’s experiment was that standard CRISPR-Cas9 creates a full double-strand break in DNA, which the cell must then repair. That repair process is inherently error-prone, especially when the goal is a precise change rather than simply knocking out a gene. Newer approaches try to avoid the double-strand break altogether.
Base editing, developed in the years surrounding the CRISPR-baby episode, can chemically convert one DNA letter to another without cutting both strands. Two classes exist: cytosine base editors, which convert C·G pairs to T·A pairs, and adenine base editors, which do the reverse. Prime editing goes further, able to make any of the twelve possible single-letter swaps as well as small insertions or deletions, all without a full double-strand break.19PubMed Central. CRISPR-Cas9 DNA Base-Editing and Prime-Editing
Comparisons in cardiac-disease models found that base editing achieved the highest editing efficiency and protein expression, while prime editing was better at minimizing off-target mutations and maintaining cell viability.20Indus Journal of Bioscience Research. A Comparative Analysis of CRISPR-Cas9, Base Editing and Prime Editing Technologies for Precision Gene Therapy in Cardiac Disease Applications These tools are not perfect — base editors can still produce off-target structural variations in embryos — but they represent a meaningful step toward precision. Whether they will eventually make germline editing safe enough to attempt responsibly is a question that most scientists consider premature, given how far the entire field still needs to advance.
The Disability Rights Perspective
The ethics conversation around gene editing tends to focus on safety and consent, but disability rights advocates raise a different category of concern entirely. If gene editing can eliminate genetic conditions before birth, what does that say about people living with those conditions now? Some scholars and community members argue that disabilities are not simply problems to be deleted from the human experience. The social model of disability holds that people experience their differences as disabilities in large part because of the world around them, not because of inherent deficiency.
The Deaf community offers a vivid example. Many Deaf people view themselves as part of a distinct culture with its own language and social norms, rather than as “disabled.” Editing embryos to confer the ability to hear might be seen by hearing parents as a gift, but some in the Deaf community view it as an erasure of their identity and culture.21PubMed Central. The Ethics of Human Embryo Editing via CRISPR-Cas9 Technology: A Systematic Review of Ethical Arguments, Reasons, and Concerns Beyond individual identity, the broader worry is that widespread use of gene editing to eliminate disabilities could increase intolerance toward people with disabilities, reduce society’s willingness to fund accommodations and support, and further stigmatize families who have a child with a genetic condition, whether by choice or by chance.22Journal of Engineering, Science and Technological Trends. The Human Rights Implications of Scientific Progress: A Case Study on Gene Editing and Disability Rights
These concerns do not necessarily lead to a blanket objection to gene editing. Many disability rights advocates draw their own line between conditions that cause severe suffering and those that are primarily disabling because of social barriers. But their perspective highlights something that a purely medical or technical framing misses: the decision about which traits count as disorders worth eliminating is not a neutral scientific judgment. It is a social and political one, shaped by who gets to make it and whose experience is centered in the conversation.
Where CRISPR Came From
The technology that made all of this possible started with bacteria. CRISPR-Cas systems evolved as a form of immune defense in bacteria and archaea, protecting them against viruses.23PubMed Central. CRISPR-Cas9: A History of Its Discovery and Ethical Considerations of Its Use in Genome Editing When a bacterium survives a viral attack, it stores a snippet of the virus’s DNA in its own genome, within a CRISPR array. If the same virus attacks again, the bacterium uses that stored snippet as a guide to recognize and cut the invading DNA. It is, in essence, a molecular memory system for past infections.24PubMed Central. Casposons – silent heroes of the CRISPR-Cas systems evolutionary history
Researchers realized they could repurpose this defense system by designing their own guide RNAs, directing the Cas9 enzyme to cut virtually any DNA sequence of their choosing. The breakthrough paper came in 2012, and within a few years the technology had spread to laboratories around the world. The speed of adoption was staggering, but the ethical frameworks for its most consequential potential application, editing the human germline, lagged far behind. Viruses, for their part, developed counter-strategies of their own: anti-CRISPR proteins that can shut down the bacterial defense system, a reminder that even in nature, this technology operates within an evolutionary arms race.25PubMed. The Anti-CRISPR Story: A Battle for Survival That dynamic, powerful tools that outpace the systems meant to govern them, turned out to be a reasonable metaphor for the human application as well.