For two decades, the lab-grown mini-organ has been a useful fraud. You can
coax a clump of human stem cells into something that looks like a kidney, a
liver, a piece of cerebral cortex. Slice it, stain it, photograph it: a
recognizable organ in miniature. Try to keep it alive past about four
millimeters across, and it dies in the middle. No blood vessels means no
oxygen past a few cell layers. No nerves means it doesn’t know what an
intact organ is supposed to feel. No immune cells means a brain organoid is
a brain without microglia, which is to say, not a brain. For most of the
2010s, “organoid” was shorthand for interesting blob in a dish.

The blob is finishing.

In the four months since January, the US Patent and Trademark Office has
granted 24 utility patents whose abstracts contain the word “organoid.” The
2025 full-year total was 51. In 2017 it was five. The grants themselves are
small and technical and read like service manuals for the missing parts: a
soft, stretchable microelectrode array with a “3D pocket” sized to cradle an
organoid (US 12,533,673). A method for inducing functional blood vessels
inside a beta-cell organoid using a transcription factor called ETV2
(US 12,576,114). A way to drop primitive macrophages onto a 15-to-30-day-old
brain organoid so they mature into microglia and the brain finally has its
immune system (US 12,577,531). A 3D-bioprinted cerebral cortex organoid
plumbed into a microfluidic chip that simulates cerebrospinal fluid
circulation (US 12,571,792). And, granted in mid-April, US 12,600,943, which
teaches you how to wire a Hedgehog signaling pathway into a precursor cell so
it becomes a neural crest cell, the kind your gut and your skin grow nerves
out of, the building block for an “innervated organoid.”

Mentally cross out the word “organoid” in those five claims. You are still
left with five different ways to give a piece of artificial tissue something
it didn’t have before: blood vessels, nerves, immune cells, an electrical
interface to the outside world. That is not a vocabulary trend. That is a
field methodically completing the parts list.

Cincinnati, of all places

The most-cited recent grants in this run cluster around Children’s Hospital
Medical Center in Cincinnati: eight US patents involving organoids since
the start of 2022, more than Harvard, more than the Hubrecht Institute in
the Netherlands where the original intestinal organoid was made in 2009. The
hospital’s Center for Stem Cell and Organoid Research (CuSTOM) is run by
Takanori Takebe, a hepatologist who in 2013 published the first liver
organoid that grew its own blood vessels when transplanted into a mouse.
Twelve years later, in Nature Biomedical Engineering on June 25, 2025,
Takebe and collaborators reported the IMALI method, an “inverted
multilayered air–liquid interface” recipe that lets four cell types
self-organize into a 3-millimeter dome of liver tissue with internal
arterial and sinusoidal vasculature, no transplant required. When the team dropped the organoids
into hemophilia A mice, the engineered livers secreted human Factor VIII and
the bleeding stopped for five months.

That same Cincinnati group is the assignee on the new “shaped organoid” and
liver-organoid patents that anchor this trend. The pattern of inventors (Takebe,
Michael Helmrath, Maxime Mahé, Holly Poling) repeats across the filings. They are quietly converting a research program into IP that reads
like a manufacturing recipe.

Why the patent office is paying attention now

Three things changed at once. The first is regulatory. In December 2022
Congress passed the FDA Modernization Act 2.0, which struck the
animal-testing requirement out of the federal statute that has governed new
drug applications since 1938 and explicitly named “cell-based assays,
microphysiological systems, and bioprinted or computer models” as permitted
substitutes. On April 10, 2025, the FDA followed up with a Roadmap to
Reducing Animal Testing in Preclinical Safety Studies, telling sponsors of
monoclonal antibodies that organ-on-chip and organoid data could replace
some animal toxicology work and projecting that animal testing should become
“the exception rather than the norm” within three to five years. In December
2025 the Senate passed the Modernization Act 3.0 by unanimous consent,
adding a formal qualification process for the new methods. A regulator
saying “we will accept this evidence” is the only thing pharma needs to
hear.

The second is economic. AxoSim, which grows neural “microbrains,” recently
ran a Takeda-collaborated test of 84 known drugs in which the microbrains
correctly flagged neurotoxic compounds with high specificity. That is a
benchmark the organoid field has been chasing for a decade. With
Modernization Act 2.0 in force, every replicable benchmark study is a sales
call.

The third is computing. On March 2, 2025, an Australian startup called
Cortical Labs unveiled the CL1 at Mobile World Congress in Barcelona: a
$35,000 rack-mounted unit that sits 800,000 lab-grown human neurons on top
of a high-density multielectrode array and rents the result over the
internet as “Wetware as a Service.” The CL1 is the descendant of Cortical’s
2022 demonstration in which a sheet of cultured cortical neurons learned to
play Pong, and it consumes between 850 and 1,000 watts for a full rack of
30 units. Thomas Hartung at Johns Hopkins, who coined “organoid
intelligence” in a 2023 Frontiers paper that is up to 243 citations and
counting, has an NSF grant to build organoids that can play simple video
games and steer small robots. Whether or not biocomputers ever displace
silicon, and reasonable people are skeptical, the people building them need
exactly the same enabling technology that drug-screening organoids need:
vasculature so the tissue doesn’t die mid-experiment, microglia so it
doesn’t quietly inflame, and electrode arrays in the right geometry. The
patents being granted now are infrastructure for both customers.

The literature has been here for a while

Organoid is no longer a niche search term. OpenAlex, the open citation
graph, indexes 7,345 scholarly works published in 2025 with “organoid” in
title or abstract, up from 55 in 2010, a 133-fold increase. The 2024 paper
“A microfluidic platform integrating functional vascularized organoids-on-chip”
already has 224 citations. “Vascularized human cortical organoids
(vOrganoids),” published in 2020, has 364. The field’s leading review on
organoid intelligence has been cited 243 times in three years. The papers
arrived first; the patents are arriving now, which is the order things
usually arrive in when an academic technique gets close to a market.

What the patents tell you that the headlines don’t

If you are trying to decide where this is going, ignore the Nature
press releases for a moment and read the claims. The 3D pocket in
US 12,533,673 is sized for a specific organoid diameter and the elastomer
substrate is specified to within a stiffness range. US 12,576,114 names
ETV2, a single transcription factor, as the switch that turns endothelial
cells into pancreatic-islet-grade vasculature. US 12,571,792’s hydrogel
recipe (gelatin, alginate, hyaluronic acid) is the kind of detail you
include when you want to stop someone else from selling cerebral-cortex
chips. These are not aspirational research disclosures. They are the
boring parts you patent when you intend to ship.

The blob in a dish is being broken into modules and the modules are being
fenced. A pharma company that wants a vascularized human liver model in
five years will not be deciding whether to use organoids; it will be
deciding which patent stack to license. The same is true, eventually, for
whoever wants to rent compute that runs on cortex.

A blob is an interesting curiosity. A blob with blood, nerves, an immune
system, and an I/O port is something else. The 24 grants from this
year so far are the patent office quietly signing off on the second one.

Method note: counts were drawn from the USPTO bulk grant feed for utility
grants between 1 January 2014 and 28 April 2026, filtered by occurrence of
“organoid” in title or abstract using full-text search. Recent assignees
(2022–) were tallied directly from USPTO assignment data; “Children’s
Hospital Medical Center” is the formal name of Cincinnati Children’s. The
literature counts come from OpenAlex (357M scholarly works), filtered the
same way. External claims about FDA actions, the Cortical Labs CL1 launch,
and the Takebe IMALI paper are sourced from the FDA’s published roadmap,
Mobile World Congress 2025 coverage, and Nature Biomedical Engineering;
inline links would land on the relevant primary sources. Patent grant counts
move slightly as USPTO updates trickle in; the figures here reflect the
state of the corpus on 3 May 2026.