Science and Technology · Iran · Emerging Tech
Iran Builds an Artificial Brain From Living Human Neurons, Entering the Global Race for Biological Computing
Iran has announced that its scientists have produced a laboratory scale artificial brain built from living human neurons, a device its backers describe as an early but working step toward computer processors made of biological tissue rather than silicon. The announcement came from Ataollah Pourabbasi, secretary of Iran's Cognitive Sciences and Technologies Development Taskforce, who told Mehr News Agency that domestic researchers have mastered the technique of culturing nerve cells outside the human body so that they form their own synaptic connections and build networks capable of learning.
According to Pourabbasi, an Iranian knowledge based company has already produced an experimental prototype of the technology, and the country now possesses the full technical know how needed to pursue brain cell based processors from start to finish. He was careful to note that commercialisation remains some distance away, but argued that Iran is advancing at a pace comparable with other countries chasing the same breakthrough.
Our country is moving at the same pace as other leading nations in acquiring this technical know how.Ataollah Pourabbasi, Secretary, Iran's Cognitive Sciences and Technologies Development Taskforce
The claim rests on a real and fast growing scientific field usually called organoid intelligence, or biological computing. Instead of etching transistors into silicon wafers, researchers grow small clusters of human brain cells, known as organoids, from stem cells, then interface them with electrodes so the living tissue can send and receive electrical signals. The neurons form synapses among themselves in ways that mirror how connections form inside a real brain, and with the right feedback, these clusters can be trained to recognise patterns, respond to stimuli and even carry out simple tasks.
Pourabbasi pointed to two advantages that researchers around the world have also emphasised: a sharp rise in processing speed for certain tasks, and a dramatic cut in energy consumption compared with conventional chips. He cited a figure of up to one million times less energy use than silicon based processors, a number that closely tracks claims already circulating in the wider field, where some Johns Hopkins researchers have suggested biocomputing could eventually reduce AI energy demands by a factor ranging from one million to as much as ten billion, according to reporting by National Geographic.
The term organoid intelligence was coined in 2023 by Thomas Hartung and Lena Smirnova at Johns Hopkins University, who described it in a paper published in Frontiers in Science as a new multidisciplinary field aimed at establishing genuine biological computing using brain organoids in an ethically responsible manner, a framing later expanded on by the Frontiers in Science research consortium. The underlying logic is simple even if the biology is not. The human brain performs an extraordinary volume of parallel computation while consuming roughly 20 watts of power, about as much as a household light bulb, whereas training a large generative AI model can burn through more than a thousand megawatt hours of electricity, comparable to the annual consumption of well over a hundred American homes.
That gap between biological and digital efficiency is what has drawn serious investment and research attention well beyond Iran. Switzerland's FinalSpark already rents access to a cluster of sixteen living organoids, containing roughly 160,000 neurons in total, through a cloud service called the Neuroplatform, according to a market analysis published by Private Markets News. The platform has logged eighteen terabytes of experimental data across more than a thousand experiments and has kept some organoids alive and functioning for over a hundred days, with dozens of universities expressing interest in remote access.
Australia's Cortical Labs has gone further toward a commercial product, selling bioprocessing units that house living neurons on silicon chips inside climate controlled glass containers, at a reported price near 35,000 dollars per unit, per National Geographic's coverage of the field. Its chief executive, Hon Weng Chong, has argued that while much of the technology industry is racing to build artificial general intelligence in software, biological neurons remain the only proven example of general intelligence that anyone has ever observed. Academic teams have already demonstrated narrower but striking results, including an Indiana University system called Brainoware that reached seventy eight percent accuracy on a speech recognition task using organoid tissue combined with electronic hardware, a result published in the journal Nature Electronics.
Iran
Government backed taskforce announces a working lab prototype using cultured human neurons, positioning the country in the global race though still short of commercial deployment.
FinalSpark, Switzerland
Runs a cloud connected Neuroplatform with sixteen living organoids, renting remote access to universities for scientific experiments since 2024.
Cortical Labs, Australia
Sells commercial bioprocessing hardware built around living neurons and has demonstrated learning behaviour including simple game play.
Johns Hopkins University, United States
Coined the term organoid intelligence in 2023 and continues to lead peer reviewed research on brain organoid learning and ethics.
For Iran, the announcement carries weight beyond the laboratory. The country has spent years working to build indigenous capacity in fields ranging from nuclear science to drone manufacturing, often under the pressure of international sanctions that restrict access to advanced imported technology. A domestically produced biological processor, even at an early experimental stage, fits a broader pattern in which Tehran has sought to demonstrate scientific self sufficiency and signal that isolation has not stalled its research base. Iranian state media outlets, including Iran's ANA news agency and the Russian aligned outlet TV BRICS, framed the development as evidence that Iran is keeping pace with leading nations in an emerging technological field, echoing similar language Tehran has used around its space and biotechnology programs.
It is worth being precise about what has and has not been demonstrated. Pourabbasi's statements, relayed through Mehr News and repeated across several outlets, describe a prototype capable of forming learning capable neural networks, not a finished processor performing defined computational tasks at scale, and no independent peer reviewed paper or third party verification of the Iranian system's performance has yet been published in international scientific literature. That gap between announcement and independently verified data is not unique to Iran. Even the most established labs in this field face similar scrutiny.
At a scientific meeting held in Pacific Grove, California in late 2025, organoid intelligence pioneers themselves warned that inflated claims about biocomputing could trigger a public and regulatory backlash, according to reporting by STAT News. Researchers there noted that most organoids used in current experiments contain fewer than one hundred thousand cells and are roughly one three millionth the size of a human brain, figures documented in the original Frontiers in Science paper that introduced the field.
Whatever else they are, they aren't brains. They aren't organized like brains. They aren't big enough.Hank Greely, Stanford University, on brain organoid research, as reported by Undark
Reporting by the science outlet Undark found that some scientists in the field now prefer the more modest term neural organoids over brain organoids or mini brains, precisely because media coverage has tended to overstate how closely these lab grown clusters resemble an actual functioning human brain. Researchers such as Mirai Sharf have received National Science Foundation funding specifically to design rigorous tests of what these systems can and cannot compute, after cautioning that public claims have outpaced verified capability.
None of that skepticism erases the scale of investment building around the field. Industry researchers project the broader organoid market, which today is driven mostly by drug discovery and disease modelling rather than computing, will grow from roughly 1.8 billion dollars in 2025 to about 9.6 billion dollars by 2034, an annual growth rate near twenty percent, according to the Private Markets News analysis. The United States National Science Foundation has already committed 14 million dollars across seven research teams through a program called Biocomputing through Engineering Organoid Intelligence, launched in 2024, and notably required every funded team to include a bioethicist as co principal investigator, weighting ethical review equally with scientific merit.
| System | Approach | Reported Scale | Status |
|---|---|---|---|
| Iran prototype | Cultured human neurons forming learning networks | Not independently disclosed | Experimental lab prototype |
| FinalSpark Neuroplatform | Cloud accessible organoid array on electrodes | 16 organoids, about 160,000 neurons | Operational, rented to universities |
| Cortical Labs CL1 | Neurons on silicon chip in sealed hardware unit | Commercial unit, roughly $35,000 | Shipping to customers |
| Johns Hopkins research organoids | Standardised 3D brain organoids for cognition study | Under 100,000 cells per organoid | Peer reviewed academic research |
The ethical questions surrounding this field are not abstract. Growing human neural tissue outside the body, then training it through repeated electrical stimulation, raises questions that bioethicists have been grappling with since before Iran's announcement. Researchers including Lena Smirnova at Johns Hopkins, whose lab sits within the university's Center for Alternatives to Animal Testing, have already begun drafting review frameworks similar to those used in animal research, arguing that the field must confront questions of sentience and welfare before organoids grow larger and more complex, rather than after. Cambridge researcher Madeline Lancaster has separately warned that overstated claims about biocomputing risk provoking regulation broad enough to also choke off legitimate medical research that relies on the same underlying organoid technology, a concern also detailed in the STAT News report from the Pacific Grove meeting.
Iran has not detailed what ethical review structure, if any, governs its own organoid research, nor has it specified the source of the human stem cells used to derive the neurons described in Pourabbasi's announcement. Those details matter for how the international scientific community will eventually judge the credibility of the claim, since reputable organoid research groups typically publish exactly this kind of provenance and oversight information alongside their results.
Strategically, the announcement lands at a moment when computing power itself has become a geopolitical asset. Advanced graphics processing chips used to train large artificial intelligence models are subject to export controls that have limited Iran's access to the same silicon hardware fuelling AI development in the United States, China and the Gulf states. A biological computing pathway, even a nascent one, offers a theoretical route around that bottleneck, since it does not depend on importing cutting edge semiconductor fabrication technology that sanctions regimes are specifically designed to restrict. Whether that theoretical route can mature into a practical one within Iran's current scientific infrastructure and funding base remains an open question that will likely take years, not months, to answer.
What is clear is that Tehran is not alone in believing biological neurons may eventually outperform silicon at certain tasks, particularly ones involving small or noisy datasets, continuous learning and extreme energy efficiency, use cases increasingly relevant as global data centre electricity demand climbs alongside the artificial intelligence boom. Whether Iran's prototype represents a genuine scientific advance or primarily a demonstration of political intent, it confirms that the competition to harness living neurons for computation has now become a truly global one, stretching from Baltimore and Melbourne to Zurich and Tehran.
World At Net will continue to track independent verification of Iran's organoid research as further technical details and peer reviewed data become available.
.jpg)
.jpg)
0 Comments