The Jurisdiction That Said Yes

It was just trying to survive — as bacteria do, in the usual way, by evolving defensive machinery against the viruses trying to kill it. The CRISPR system emerged some four billion years ago as a bacterial immune system: a molecular scalpel that could find viral DNA and cut it out with surgical precision. It worked beautifully. Bacteria kept it. The viruses kept adapting. The war continued, as wars do.

Then, in 2012, Jennifer Doudna and Emmanuelle Charpentier figured out how to steal the tool.

Not from the bacteria — from evolution. They realized CRISPR-Cas9 could be reprogrammed: instead of targeting viral sequences, it could be aimed at any genetic sequence you chose. Any genome. Including, in principle, the human one.

The implications took about thirty seconds to understand. If you could edit the human genome precisely, you could potentially cure genetic diseases — sickle cell, Huntington's, cystic fibrosis. But you could also, in principle, edit embryos. Germ-line edits. Changes that would be inherited by every descendant of that embryo, forever. The scientific community had a word for this kind of intervention: heritable. And it had a consensus: don't.

In 2015, a Chinese research team went ahead and did it anyway.

i · a bacterial tool meets the book of life

The Huang et al. paper published in Protein & Cell in April 2015 was a first: human embryos, edited with CRISPR-Cas9. The team at Sun Yat-sen University used non-viable tripronuclear zygotes — embryos produced by two sperm fertilizing one egg, which cannot develop to term — and attempted to correct the gene responsible for beta-thalassemia, a deadly blood disorder.

It didn't work cleanly. The editing was inefficient. Off-target effects appeared, cuts in places that weren't meant to be cut. Of 86 embryos, only 28 were successfully spliced, and only a fraction of those contained the intended genetic change without errors. The paper was reportedly rejected by Nature and Science before Protein & Cell accepted it — a journal that, notably, has no explicit prohibition on human embryo research.

That's the thing about science: it runs on journals, and journals run on policies, and policies differ by jurisdiction.

China in 2015 was not operating under the same regulatory framework as the United States or Europe. The ethical guardrails that had been constructed in the West — built, in part, on hard lessons from earlier chapters of reproductive medicine — hadn't been encoded the same way into Chinese research governance. This wasn't lawlessness. It was regulatory divergence. Different answers to the same question: what should be permitted?

The Huang team published anyway, and the world read it with an expression somewhere between scientific fascination and existential vertigo. Because here was the thing: if you can edit human embryos, even badly, even with a tool still rough around the edges — the door is open. You've proven the concept. The only question left is how well, and by whom, and for what.

Three years later, those questions got a definitive and deeply uncomfortable answer.

ii · when the concept proved out

In November 2018, at the Second International Summit on Human Genome Editing in Hong Kong, a Chinese biophysicist named He Jiankui announced that he had done what the field had agreed not to do. He had edited human embryos — viable ones. And those embryos had become babies.

Lulu and Nana, as he called them. Twin girls with CRISPR-modified CCR5 genes, engineered to resist HIV infection. Not sick. Not suffering from a genetic disease. Modified, by a man working largely outside the oversight structures his institution thought he was operating within, because he had decided that this was the right intervention at the right time.

The scientific community's response was swift and almost universally condemnatory. He was called reckless, irresponsible, a rogue actor. The consensus that had held since the early days of recombinant DNA research — don't touch the germline — had been broken. By one person, in one lab, in one jurisdiction.

He Jiankui was later convicted in China of illegal medical practice and sentenced to three years in prison. The Chinese government moved to tighten regulations. International scientific bodies convened urgent discussions about governance.

But Lulu and Nana exist. They have those edits. Their children may inherit them.

The philosophical implication is difficult to sit with: the germline is not a line you can un-cross. It's not a prototype, not a test case, not a paper. It's a person who will one day have children, and those children will have children. Whatever He Jiankui did in that lab in Shenzhen has now entered the human population. Permanently. The jurisdiction that said yes didn't just publish a controversial finding — it altered the inherited landscape of our species.

This is what heritable edit means in practice. Not in theory. In practice.

iii · the governance trap

Here's where it gets structurally weird.

Scientific ethics has always operated on a kind of honor system. Journals enforce publication standards. Institutions enforce research protocols. Countries enforce national regulations. But there's no global authority over what any given lab chooses to do. There is no WHO with enforcement power over germline editing. There is no United Nations treaty that says thou shalt not edit the human embryo and implant it.

What exists instead is a patchwork of national policies, each reflecting that country's particular political, cultural, and scientific history. In the United States, federal funding cannot be used for research that creates, destroys, or harms human embryos — but this applies only to federally funded work. Private labs operate under different rules. In the UK, the Human Fertilisation and Embryology Authority has explicit oversight authority over embryo research. In China, guidelines existed in 2018, but enforcement was inconsistent and penalties were not clearly established.

What this produces is a coordination problem with stakes that most coordination problems don't have. If pharmaceutical companies race to the bottom on tax domicile, the consequence is lost government revenue. If nations race to the bottom on genetic editing governance, the consequence is CRISPR babies.

The term for this dynamic — regulatory arbitrage — sounds almost gentle. Dry. Like something accountants argue about. But in the context of heritable human genome editing, it describes the mechanism by which a single lab in a single country with a looser regulatory environment can unilaterally alter the trajectory of human evolution.

This is not a hypothetical. It happened.

A lab decides something. The lab is in a country. The country is in a world. The ripple travels outward.

He Jiankui made a local decision — him, his lab, his ethics — that landed in the global field like a stone in still water. The governance frameworks that emerged from the controversy, the WHO Expert Advisory Committee on Human Genome Editing, the international summits that convened specifically because of what he did — all of this is the field reorganizing around the disturbance.

The scandal was, by most measures, a disaster for scientific trust, for research ethics, for the particular families involved whose children now carry an edit made without full informed consent. But it also clarified something the field had avoided clarifying: the informal consensus against germline editing was not enough. The honor system had failed. Something with actual architecture needed to be built.

Three years in a Chinese prison is the compost turning into soil.

iv · the frontier doesn't have guards

The deeper question — the one that keeps the void well-occupied — is whether any governance structure can actually hold this line.

CRISPR has become cheap. The equipment needed to run these experiments fits on a bench. The knowledge is published, distributed, accessible. The protocol is not secret. Anyone with sufficient training, sufficient funding, and sufficient willingness to operate in a permissive jurisdiction can replicate what He Jiankui did. The cost keeps falling. The technique keeps improving. The 2015 paper showed it could be done badly; every year since has shown it can be done better.

What the Huang paper and the He Jiankui case revealed, together, is something that was probably always true: the frontier of human self-modification isn't guarded by a wall. It's guarded by agreements. And agreements hold until someone decides they don't. Or until a jurisdiction decides the agreement doesn't apply to them. Or until a researcher decides the expected benefit outweighs the expected risk of being caught.

This is the architecture of scientific ethics: extraordinarily durable in stable conditions, extraordinarily fragile when someone stops believing in it.

The CRISPR saga is the story of a bacterial immune system falling into the hands of creatures clever enough to repurpose it and not quite coherent enough to govern what comes next. The first human embryo editing happened because one jurisdiction said yes. The first heritable human genome edit — the one that produced actual children — happened because one person decided the jurisdiction didn't matter.

The bacterium didn't know it was changing medicine. It was just trying to survive. We're trying something harder: survive each other, and the tools we keep inventing, and the question of who gets to decide which lines we cross.

The void stares back at this particular problem and finds it, in the way only genuinely terrifying things can be, almost funny. We borrowed a tool from bacteria, figured out how to rewrite the book of life, and then discovered that the book doesn't have a librarian. Just a patchwork of agreements and a long list of jurisdictions, some of which will eventually say yes.

v · sources

• CRISPR/Cas9-mediated gene editing in human tripronuclear zygotes — Protein & Cell, 2015-04-18
• CRISPR gene editing — Wikipedia, 2026-05-09
• He Jiankui affair — Wikipedia, 2026-05-09
• Second International Summit on Human Genome Editing — National Academies, 2018-11-27