Two technologies intervene on the same underlying biological substrate — human biological material. One is treated like a constitutional emergency. The other got a product launch at a tech conference.
Here's a puzzle that doesn't get talked about enough.
If you edit the human germline in a way that can be inherited by future generations, you run straight into some of the strictest law on the books. For European states bound by the Oviedo Convention, Article 13 places one of the strongest legal restrictions on it: interventions intended to introduce modifications into the genome of descendants are prohibited. This is treated as close to a sacred line.
Meanwhile, in March 2025, a company called Cortical Labs stood up at Mobile World Congress in Barcelona and launched CL1 — the world's first commercially available "biological computer." It's built on real, living human neurons, grown in a dish, wired into a feedback loop, and rented out over the cloud as "Wetware-as-a-Service." You can, today, pay to run compute jobs on a silicon-based device carrying living human neurons.
One of these things runs straight into a hard legal prohibition. The other gets a keynote.
Why?
The obvious explanation is "consciousness." Maybe we're fine with organoid computers because we're sure they're not conscious, and we're terrified of germline editing because it touches something more sacred than a mere nervous system.
But that doesn't hold up. Regulators openly admit they don't know whether lab-grown brain tissue can have some kind of experience. ISSCR, the leading international professional society issuing stem-cell research guidelines, actually discussed this directly in its 2021 guidelines — should organoids that show signs of complex neural activity get special ethical scrutiny? Their answer, in effect, was: we currently have no biological evidence of consciousness or pain perception in these systems, so routine CNS-organoid research generally remained within the lightest oversight tier, Category 1A. That's a statement about the current state of evidence, not a confident verdict on the metaphysics of machine or organoid minds.
So the law isn't actually resolving the hard problem of consciousness before deciding how to regulate these things. It's doing something else. And once you see what that something else is, the asymmetry stops looking like a mistake and starts looking like a pattern.
These are not personhood criteria. They are reconstructed filters explaining when subject-status recognition becomes legally available in the first place.
Here's the argument: courts and regulators don't need to resolve consciousness before their existing recognition structures produce a classification. No court has ever laid out a two-part "lineage and intent" test by name — nobody's citing this. What's being claimed here is narrower: two much older, much cruder legal categories — who a thing is attributed to, and whose purposes are actually steering it — already do most of the classificatory work before anyone gets around to asking about consciousness. Once you look for that pattern across these two cases, it's hard to unsee.
Trip-wire one: Does this thing have a lineage?
Legal systems routinely organize important legal relationships around ancestry and descent — parents, children, inheritance, and generations. That's not a philosophical flourish; it's baked into how property, inheritance, and personal identity work.
Germline editing hits this trip-wire directly. It intervenes exactly at the point where one generation hands biological material to the next — which is the lineage chain itself. That's a big part of why the reaction is so severe: this isn't just "editing a person," it's editing the chain.
Organoid computing does not ordinarily activate this particular lineage-to-personhood pathway — and, in existing law, excised biological material is generally handled through other categories, such as consent, provenance, and ownership, rather than through the donor's lineage.
The story of HeLa cells shows what this severance looks like outside the courtroom. In 1951, cells taken from Henrietta Lacks during cancer treatment at Johns Hopkins were developed into the HeLa cell line without her knowledge or consent, and they went on to become one of the most important materials in modern biomedical research. The legal claims of Lacks and her descendants did not simply travel with the cells: in 2023, after decades in which the family received no share of the commercial returns from the cell line, her family's claims against Thermo Fisher Scientific ended in a confidential settlement — not a judicial ruling that the family owned the cell line. HeLa isn't a legal precedent for the argument here. It's something more useful: a real-world demonstration of how far biological material can travel once the law stops treating where it came from as a reason to keep a claim alive.
Back in 1990, in a case called Moore v. Regents, the California Supreme Court drew a narrower line than people often remember — and it's the narrower version that matters here. John Moore sued after doctors used cells from his removed spleen, without properly telling him, to build a hugely valuable, patented cell line. He won on the nondisclosure: the court agreed his doctor should have told him what he planned to do with the tissue. But on the separate question of whether Moore still owned the cells once they'd left his body, the court rejected his claimed continuing property interest in the excised cells and the patented cell line — a specific conversion claim, not a categorical ruling that no legal interest of any kind can survive excision.
That second half of the ruling is the one with quiet, long-running consequences. It didn't sever anyone's lineage — that was never what was on trial. What it did was close off one specific legal pathway: a conversion claim resting on biological origin alone doesn't survive tissue being excised and put to new use. So when a research lab grows neurons into an organoid and wires it into a computer, nobody has to ask a family-tree question to get to a property answer — one property-law pathway underneath that question had already been closed off thirty-five years ago, in an unrelated case about a spleen, not because any court examined this tissue's ancestry and rejected it. The remaining legal questions around organoid tissue — who consented to its use, who owns the resulting product, what regulatory oversight applies — are real, and they're routed instead into provenance, consent, and ownership review, not into a lineage inquiry.
Trip-wire two: Was the entity's purpose externally assigned at the point of its creation?
The second check isn't "was this influenced by someone else" — almost everything is; that test would catch every child anyone ever raised. It's narrower than that: was there a point at which this entity's purpose was assigned entirely by someone else, before it had any capacity to do otherwise? Philosopher Harry Frankfurt argued that being an autonomous agent isn't about having desires that appeared out of nowhere — it's about having some ongoing capacity to step back and reflect on your own desires, to potentially want something different than what you currently want. Call it a "reflective gap." (No court cites Frankfurt by name here — I'm borrowing his vocabulary as an explanatory lens for naming a distinction, not claiming judges apply it as legal doctrine.)
You've probably already met a version of this question, even if you didn't file it under "legal theory." It's the premise of My Sister's Keeper, which was loosely inspired by a real-life case that became public in 1990, in which a couple named Abe and Mary Ayala conceived a child specifically to be a bone marrow donor for their teenage daughter's leukemia. The novel's version of that child eventually sues her parents for the right to say no to a further donation. Strip away the courtroom drama and what's left is trip-wire two in miniature: not "was this child wanted" or "was this child loved," but "was this person's purpose assigned entirely by someone else, before they existed to have any say in it?" That's the same question, at a much colder register, that decides whether a dish of neurons gets treated as a subject or as a product.
Two different things are true about germline editing, and it's worth keeping them separate. The edit itself is exactly the kind of event this trip-wire is built to catch: a design decision imposed on someone before they exist to object to it, made entirely by someone else's intent. That's what trips the wire. What happens afterward is a different question — and there, a person who's been genetically edited still goes through an entire, messy, unpredictable life after that edit: childhood, environment, chance, biology doing its own unplanned thing. The edit changes one starting condition; it doesn't lock down everything that follows. The reflective gap survives in the life that comes after. But the wire already tripped at the moment of imposition, before any of that unpredictability had a chance to run.
An organoid computer is built differently, on purpose. Neurons in a system like CL1 sit inside a tightly closed loop: electrodes stimulate them, their activity gets read out, and a training signal — designed entirely by engineers — tells the system what counts as a "good" or "bad" outcome. Interestingly, this design lineage goes back to the free-energy principle, a genuine and influential theory in neuroscience about how brains minimize "surprise." The original 2022 DishBrain experiment, where neurons in a dish learned to play Pong, was explicitly motivated by that theoretical picture — and there really is some genuine self-organizing, adaptive behavior happening at that level.
But notice what the engineers still control, no matter how the internal algorithms evolve: they set what counts as a win. The neurons may improvise tactically toward whatever goal they're given, but the architecture provides no legally recognizable channel through which that goal itself — the strategic objective — could be revised from within. That's the second trip-wire failing to trip: not because there's zero biological unpredictability in the system, but because the system-level objective remains externally specified, whatever adaptive behavior occurs beneath it.
Lay it out as a simple grid and the whole asymmetry falls into place. On this reconstruction:
Nobody had to solve consciousness to get this outcome. The legal system doesn't need to know what, if anything, it's like to be a dish of neurons hooked up to a cloud API. It just needs to check two much older, much cruder things: does this have a family, and does it have any purposes of its own that aren't someone else's. Fail both checks, and there's little reason for the law to open a personhood inquiry at all — the material just stays, by default rather than by declaration, in the ordinary categories of property, research regulation, and consent.
One clarification, since it's the question this audience will reach for immediately: none of this is a claim about whether organoid computers actually deserve moral consideration on welfare grounds, independent of what the law currently recognizes. The argument here is diagnostic — it explains why the legal system reaches a stable classification without ever resolving that question, not that the classification is therefore the right one. If it turns out CL1-style systems have morally relevant experience, that's a separate finding, and the two trip-wires wouldn't automatically track it either way.
This same split shows up anywhere biological material leaves a body and enters commerce — not as one uniform "human tissue" category, but as a scatter of separately regulated paths that don't track any single logic like sentience or biological complexity.
In the US, whole organs intended for transplant generally can't be sold for money — the National Organ Transplant Act makes it a federal crime, regardless of how badly a patient needs one or how willing a donor is. Blood and plasma sit right next to that prohibition without being bound by it: paid plasma donation is a large, FDA-regulated industry in the US, built on tissue that also comes out of a body and also gets processed into a commercial product. Gametes split the difference again: the UK makes it a criminal offense to pay someone directly for eggs or sperm, but permits a capped "compensation" for the inconvenience of donating — currently £985 per cycle for eggs, £45 per clinic visit for sperm.
These examples point to the same broader feature that Moore v. Regents makes visible in a different legal context: excised human biological material does not enter one uniform legal category. Different materials are routed into different regulatory regimes, shaped by the particular histories and policy concerns of those regimes rather than by a single rule derived from the material's biological nature. Wetware hasn't been routed into any of these pigeonholes yet. It hasn't been routed anywhere. CL1 simply gets sold, the way you'd sell a sensor or a server rack, because none of the existing commercial carve-outs or prohibitions were written with it in mind, and nothing yet requires that they be.
There's a body of existing scholarship — most notably by Masanori Kataoka, Tsutomu Sawai, and their collaborators — that has done serious, careful work mapping the legal issues brain organoids raise, organized around themes like consciousness, legal status, consent, and ownership. That work matters, and most of it is organized around a "does this entity qualify" question, or a "can the donor really consent" question. What I haven't found in this literature is an explicit reconstruction of the prior question: what is the legal system's screening test actually checking for, underneath all the talk about consciousness and consent?
That's the gap this piece is trying to name. Not "is the organoid a person," but: what are the two quiet trip-wires that decide, in advance, whether that question even gets asked? Once you see lineage and intent-capture as the real operative variables, a lot of otherwise-baffling regulatory behavior starts to look a lot less arbitrary, and a lot more like a machine quietly doing its job. The same underlying move shows up somewhere much stranger, too, traced through an elephant's court case, two chimpanzees raised as if they were human children, and a rented brain-computer, in a companion piece: The checklist was never built for this.
This piece draws on legal theory around personhood and autopoietic systems (Niklas Luhmann's account of how legal systems maintain internal stability), and is a companion to a longer, more technical paper on how legal systems screen claims of subjectivity more broadly. A fully cited academic version of this argument is archived on Zenodo, DOI: 10.5281/zenodo.21768340.
