Australian startup Cortical Labs develops world's first device to run code on living human neurons

Against the backdrop of global companies racing to expand data centers to support AI models, researchers are exploring new computing approaches, including the possibility of using living human cells in computing systems.

Australian startup Cortical Labs says it has developed what it describes as the world’s first device that allows users to “run code” on living human neurons. The company built a system that combines lab-grown neurons with silicon hardware, aiming to support applications across neuroscience, disease modeling, robotics and AI.

How Cortical Labs’ CL1 works

The system, named CL1, is designed to operate by culturing neurons from stem cells and placing them on chips capable of sending and receiving electrical signals.

According to Cortical Labs, a small amount of blood or a skin sample can generate an almost unlimited source of cells, which are then transformed into neurons. The company is developing bio-computing facilities in Melbourne and Singapore, where multiple systems can be deployed and accessed remotely.

Cortical Labs says CL1 enables users to interact directly with neurons by sending input electrical signals and observing cellular responses in real time. While the device still uses silicon chips, it incorporates microelectrodes that communicate with living neurons—transmitting signals and reading responses as part of the computation.

The company describes the setup as different from traditional computing because the living cells require a nutrient-rich solution environment, an approach sometimes referred to as “wetware.” Cortical Labs also says about 120 CL1 devices operate in a small data center in Melbourne.

The company’s differentiator, it says, is standardization—making the connection between cell culture and the electronic interface simpler than approaches that require elaborate laboratory setups and highly customized processes.

Potential benefits for research and computing

Cortical Labs says its integrated platform can shorten research timelines from months or years to hours or days.

The company argues that interacting with living neurons could enable more energy-efficient and flexible computing compared with current technologies. Brett J. Kagan, Chief Science Officer and Chief Operating Officer of Cortical Labs, said biology is “extremely energy-efficient” and that humans do not require enormous data. He contrasted human perception with machine learning, noting that a child may recognize an object with only a few images, while machine learning systems may require tens of thousands or hundreds of thousands of data points depending on the problem. He also said humans can handle uncertainty and noise in information.

Cortical Labs also points to research applications enabled by human-derived cells. Neurons cultured from donated samples can reflect genetic features, which the company says may help scientists study cellular responses to treatments in the lab.

At the same time, Cortical Labs acknowledges that silicon-based computers still outperform for precise calculations and high-speed processing. It also notes that current AI systems may be approaching practical limits due to growing demand for data and compute power.

Kagan said future systems are likely to combine biology and silicon to use the advantages of both, adding that “the future of computing is when we can leverage all available tools to achieve the best results.”

Debate over effectiveness and ethical concerns

Some experts say biological systems may offer advantages in energy efficiency and adaptability, but remain skeptical about whether current methods will deliver meaningful benefits over traditional silicon approaches.

Alysson R. Muotri of UC San Diego argues that using only flat neural networks may not produce significant advantages. He said more complex structures such as organoids—three-dimensional mini-brains—could offer greater potential, though he described them as still under testing.

Muotri also highlighted ethical questions that could arise depending on the complexity of the biological system. He said simple neural networks do not pose major issues yet, but that more complex brain-like structures could create a form of experience in a culture dish, potentially leading to concerns about consciousness.

Experts say these issues may require new regulations and oversight mechanisms as the technology develops.

Cortical Labs, however, argues its approach could have ethical benefits, including reducing animal testing and increasing control over the biological system. Kagan said, “We believe this is a better approach.”

Where biocomputing may fit in the AI era

Although still in early stages and subject to ongoing debate, biocomputing is increasingly viewed as a supplementary path to traditional computing architectures in the AI era. As demand for processing power and energy efficiency grows, the combination of silicon and biology is presented as a notable direction for technology development in the coming years.