Who Gets to Call an Edited Embryo Healthy?

Origin Genomics is proposing to demonstrate that editing disease-causing genes in human embryos can produce healthy babies with benefits that outweigh the risks. The available NPR summary does not name the founders, targeted conditions, regulatory jurisdiction, editing method, or evidence already produced. It also does not report an approved clinical trial or planned pregnancy. For a project built around proof, those details must come before the promise.

The most compelling case is narrow and serious: preventing a severe inherited disease could spare a child pain, repeated treatment, disability, or early death. That aim should not be lazily collapsed into designer babies. Disease prevention and enhancement raise overlapping questions, but they are not the same proposal and should not be judged as if changing a lethal mutation were merely selecting an expensive eye color.

Still, “healthy” is doing more work than a laboratory result can carry. An embryo may show the intended genetic change while also containing unintended edits, a mixture of edited and unedited cells, or consequences that do not become visible until development. Because the alteration may be heritable, an error does not remain inside one procedure. There is no product recall for a childhood.

The social code arrives beside the science. Once intervention is sold as optimization, families may be asked to experience an unedited risk as a negligent choice rather than a condition of human reproduction. Price will decide who can purchase reassurance. Insurers, clinics, investors, and prospective parents may each define an acceptable outcome differently, while the future child supplies the body on which those definitions are tested.

Embryo selection also changes the comparison. If existing reproductive methods can reduce the chance of transmitting a particular condition without editing an embryo, a company must explain why a heritable intervention adds enough benefit to justify additional uncertainty. That answer will vary by condition and family. A polished consent form cannot turn limited evidence into informed consent, especially when the person most affected cannot sign it.

Before any clinical use, Origin Genomics should identify its method, targeted diseases, preclinical results, failure rates, off-target testing, embryo follow-up, and plans for lifelong monitoring. Regulators must decide who authorizes a trial, who can stop it, who pays for medical consequences, and whether resulting data remain public when findings disappoint investors. A private company may build the tool, but it cannot privately settle what counts as sufficient evidence for an irreversible human experiment.