Scientists Are Growing Mini Human Brains In Labs

What if a tiny sample of your skin could be turned into human brain tissue? Well, scientists have actually been doing versions of this for years.

Researchers can take adult human cells and reprogram them into induced pluripotent stem cells, or iPSCs. These cells can then be guided into becoming different types of cells, including neurons. Under the right conditions, those cells can organise themselves into three-dimensional structures called brain organoids. 

Neurons are the basic working cells of the brain. They communicate with one another using electrical and chemical signals, allowing the brain to process information and control everything from movement to memory.

They are not miniature human brains as they don't have the full structure or complexity of a real brain. But they can reproduce some features of the developing human brain, making them increasingly useful for studying how our brains develop and what goes wrong in disease.

But here's where things get interesting. These organoids aren't just models of a brain, they are actual living, human-derived neural tissue, with neurons that can connect, communicate and generate electrical activity.

Why are scientists growing mini-brains?

One major reason is disease research. Scientists can create organoids from human cells and use them to study aspects of brain development and diseases that are difficult to investigate inside a living human brain.

Researchers are already using these models to study neurodevelopmental and neurological conditions, and to investigate potential treatments.

There is also another potentially big use case in drug testing.

A drug that works in an animal does not necessarily work in a human. Brain organoids give researchers another way of testing how human-derived neural tissue responds to drugs and other substances.

The idea is not necessarily to replace animal research overnight. It is to develop more human-relevant models and, in some cases, reduce the dependence on animal testing.

The US National Institutes of Health is also putting more money into human-based research. In September 2025, NIH announced $87 million for a new Standardized Organoid Modeling Center to develop more reliable and reproducible organoid models. The centre initially focuses on liver, lung, heart and intestine, with plans to expand into brain models. In March 2026, it announced more than $150 million for human-based research designed to develop and scale methods that better simulate human biology.

But the medical applications are only one part of the story.

These mini-brains are doing something

As neurons develop inside an organoid, they form networks and communicate through electrical signals.
Researchers can detect this activity and stimulate the neurons with electrical signals.

That does not mean an organoid is thinking. But it does mean these tiny pieces of human neural tissue are showing some of the basic activity we associate with a functioning brain. The neurons connect with one another, communicate through electrical signals and respond when researchers stimulate them. As these networks develop, scientists can observe increasingly organised patterns of activity.

Simply put, researchers are not just growing a collection of human cells, they are growing living neural networks and watching them develop and respond in real time.

And that is where things get really juicy. Because if living neurons can respond to information and change their activity based on what happens around them, could they also learn?

Along came Pong

In a 2022 experiment, researchers at Melbourne-based Cortical Labs connected living human and mouse neurons to an array of electrodes and put them inside a simulated version of the classic video game.
The idea was simple yet wild - could living brain cells be taught to play a game?

The system, called DishBrain, received information about what was happening in the game through electrical signals. The neurons then produced their own signals in response.

When the system performed correctly, the neural culture received predictable feedback. When it did not, it received different, more chaotic signals.

The researchers reported that the neural cultures improved their performance during the experiment. The study described this as the neurons displaying goal-directed activity and what the authors called "synthetic biological intelligence".

Living human neurons in a laboratory were basically connected to a computer and used to play Pong.

But there is an important distinction here.

This was not a miniature human brain sitting in a dish and consciously playing a video game.

DishBrain used a layer of cultured neurons connected to electrodes. It was a biological neural network, rather than a fully formed brain organoid. The scientific journal Neuron reported at the time that hundreds of thousands of human neurons growing on an electrode-covered dish had been taught to play a version of Pong.

The experiment showed that a network of living neurons could adapt its activity when it was connected to an environment that provided information and feedback.

The researchers called this learning and described the system as displaying "synthetic biological intelligence". But that does not necessarily mean the neurons understood Pong, or that they were conscious.

Are these mini-brains conscious?

The short answer is that we don't really know yet.

Scientists do not have a simple, universally accepted test that can look at a brain organoid and tell us whether it is conscious. Electrical activity does not automatically mean consciousness, neither does responding to a stimulus.

And even the fact that a neural culture appears to learn a simple task does not prove that it is having a subjective experience.
This remains an active area of scientific and ethical debate. A major 2024 review in Frontiers in Artificial Intelligence discusses consciousness, moral status, informed consent, governance and regulation around brain organoids and organoid intelligence.

And this is where the science starts running into philosophy. If an organoid ever became complex enough to have some form of experience, how would we know?

There is currently no simple test that can answer that question.

What happens if mini-brains get bigger?

Brain organoids do not have the full vascular system of a human brain. That makes it difficult for larger structures to get enough oxygen and nutrients. Researchers are therefore working on ways to improve vascularisation and keep organoids alive for longer.

The more sophisticated these systems become, the more difficult the ethical question gets. A tiny cluster of neural cells with limited activity is one thing but what if scientists eventually create a much larger, more mature and better-connected piece of human neural tissue?

At what point would it deserve some form of protection?

That is no longer purely a science-fiction question. Researchers and bioethicists are already discussing the potential moral status of brain organoids, along with issues such as informed consent, privacy and regulation.

What happens to the cells that came from you?

A July 2026 Nature commentary warned that people who donate tissue for biomedical research may not realise that their cells could eventually be used to create biological computing systems.

That means the debate goes beyond whether a lab-grown brain deserves rights. It could also be about whether the person whose cells created it has any rights over what scientists eventually make from them.

And this is where the story moves from medicine into something that sounds much closer to science fiction.
Traditional artificial intelligence (AI) uses silicon, software and artificial neural networks to reproduce some aspects of intelligence.

When AI meets living brain cells

For years, the AI race has been about teaching machines to imitate the way brains process information. Now, a small group of researchers is exploring a radically different route: using living human neural tissue itself as part of a computer.

A growing group of researchers is asking a very different question: What if you used living neural tissue itself as part of a computer?

The field is known as organoid intelligence, or OI.

In a 2023 Frontiers paper proposing the field, researchers described organoid intelligence as an emerging approach that combines brain organoids with technologies that can send information to and receive information from living neural tissue.

The idea is not simply to replace AI with living tissue. It is to explore whether biological neural networks could become another kind of computing system, potentially working alongside conventional AI and electronics.

And researchers are already building systems around this idea.

Cortical Labs has developed the CL1, which it describes as a biological computer designed to let researchers work with living neurons as a computing substrate. Its system connects living neural cells with electronics and software.

The company has also demonstrated Doom - the classic first-person shooter video game - running on a CL1.

But again, it is important to separate what exists today from what remains speculation. Scientists have not created a humanoid robot powered by a lab-grown human brain yet, neither have they created a conscious machine from an organoid.

What they have done is connect living neurons to electronic systems and show that those neurons can respond, adapt and perform limited tasks.

The much bigger question is whether biological neural tissue could eventually become a useful new form of computing.

Researchers exploring biological computing point to possible advantages of living neural systems, including their ability to adapt and learn from relatively little information. But this remains a young field, and today's systems are nowhere close to matching the flexibility and complexity of a human brain.

For now, the idea of a robot walking around with a lab-grown human brain belongs firmly in the future. But the basic building blocks are seemingly being put in place.

What's really interesting is as the AI industry spends billions trying to build intelligence in machines, biologists are growing the raw material of intelligence in dishes.

We are perhaps still very far from creating a conscious machine this way. But for the first time, the question is no longer purely hypothetical: if intelligence can emerge from living cells, could the next generation of computers be built not just to imitate the brain, but partly from it?