Merge Labs Explained: Sam Altman’s Brain-Computer Interface Startup

Merge Labs is a brain-computer interface (BCI) startup co-founded by OpenAI CEO Sam Altman. Reports about it first came out in 2025. Its goal is to build technology that lets the human brain and computers talk to each other more directly. That idea has moved from science fiction to medical trials in the last few years. This explainer covers what Merge Labs is, how its reported approach differs from Neuralink, how BCIs work, who else is building them, and what it could all mean for everyday tech users.

A note on accuracy: Merge Labs is a young company, and many details about it have not been confirmed publicly. Where we rely on news reports, we say so. Treat any timeline or product claim with caution until the company shares it officially.

What Is Merge Labs?

Merge Labs is a research company working on brain-computer interfaces, the systems that read signals from the brain and turn them into actions on a computer. Sam Altman is one of its co-founders. Reports in 2025 described it as a new company separate from OpenAI, although the two are linked through Altman and, according to some reports, through investment interest.


The name has a history. In 2017, Altman wrote a well-known blog post called “The Merge,” where he argued that humans and machines would slowly merge over time, through phones, algorithms and, one day, more direct links. Merge Labs seems to take that idea and turn it into a company.

Here is what has been widely reported about Merge Labs so far:

  • It is focused on brain-computer interface technology.
  • Sam Altman is a co-founder, alongside other founders named in news coverage.
  • According to reports, it is looking at less invasive ways to read brain activity, possibly using ultrasound (sound waves) instead of electrodes implanted in brain tissue.
  • Product plans, release dates and detailed technical designs have not been confirmed publicly.

How Brain-Computer Interfaces Work

Your brain uses electrical and chemical signals to control everything you do. When you think about moving your hand, groups of neurons fire in a pattern. A BCI tries to capture those patterns and turn them into commands, such as moving a cursor, typing a letter or controlling a robotic arm.

Most BCIs follow the same basic steps:

  1. Sense: a device records brain activity, using electrodes, light, magnetic fields or sound waves.
  2. Process: software cleans up the signal and removes noise.
  3. Decode: machine learning models learn which patterns match which intentions, for example “move left” or “click”.
  4. Act: the decoded command controls a computer, phone, wheelchair or other device.
  5. Feedback: the user sees the result and adjusts, and the system improves over time.

AI matters here. Better models can pull useful meaning from noisy brain data, which is one reason AI leaders are interested in BCIs.

Invasive vs Non-Invasive BCIs

BCIs fall into rough groups depending on how close the sensors get to the brain:

  • Invasive: electrodes are placed inside brain tissue through surgery. These give the clearest, most detailed signals, but surgery has real risks.
  • Minimally invasive: sensors sit on the surface of the brain or are delivered through blood vessels, which avoids going deep into tissue.
  • Non-invasive: sensors stay outside the skull, such as EEG headsets. They are safe and easy to use, but the signal is weaker and less precise because the skull blocks and blurs it.

Ultrasound-based methods are interesting because researchers hope they could read brain activity in good detail without placing electrodes in the brain. This is still an active research area, and it is not yet clear how well it will work in a consumer product.

Merge Labs vs Neuralink: Different Approaches

The comparison everyone makes is with Neuralink, the BCI company co-founded by Elon Musk. Musk and Altman co-founded OpenAI together before they parted ways, so the rivalry gets extra attention.

Neuralink uses an invasive approach. A surgical robot places very thin threads with electrodes into the brain, connected to a small implant in the skull. Neuralink began human trials in 2024, and its early participants, people with paralysis, have used the implant to control computers with their thoughts.

According to reports, Merge Labs is looking at a path that doesn’t need electrodes placed deep in brain tissue, possibly using ultrasound and other methods. If that works, it could mean less risky procedures and, one day, devices more people would accept. But this is a reported direction, not a proven product. Neuralink and others already have devices in human trials, while Merge Labs’ technology has not been shown publicly in the same way.

The BCI Landscape: Merge Labs and Its Rivals

Merge Labs is joining a field that already has several serious players. Here is a simple overview based on well-known public information:

Company Approach Known Status
Neuralink Invasive: thin electrode threads placed in the brain by a surgical robot Human trials started in 2024 with people who have paralysis
Synchron Minimally invasive: a stent-like sensor delivered through blood vessels Human studies under way; users have controlled computers and Apple devices
Precision Neuroscience Minimally invasive: a very thin electrode film placed on the brain’s surface Tested in patients, mainly during brain surgeries
Merge Labs Reported to explore less invasive methods, possibly ultrasound Early stage; products and timelines not confirmed publicly

Most current BCI work is medical. It aims to help people with paralysis, ALS or speech loss communicate and use technology again. That is where the clearest benefits are, and where regulators are most likely to approve devices first.

Why Sam Altman Is Interested in BCIs

Altman has spoken for years about humans and AI growing closer. As AI systems get more capable, some technologists worry that people will fall behind if they can only talk to AI by typing or speaking. A high-bandwidth link between brain and computer could, in theory, let people work with AI much faster.

There are also practical medical reasons. Better decoding powered by AI could help people who can’t speak or move. Many experts see restoring communication as the most realistic near-term goal for any BCI company, including Merge Labs.

Ethical and Privacy Questions

Reading signals from the brain raises some of the hardest questions in tech. Before BCIs reach ordinary consumers, society will need answers to questions like these:

  • Neural privacy: who owns your brain data, and can it be sold, shared or used for ads?
  • Consent: how can users truly understand what a device records and how it is used?
  • Security: what happens if a brain-linked device is hacked?
  • Safety: what are the long-term health effects, especially for implanted devices?
  • Fairness: if BCIs boost ability, will only wealthy people get access?
  • Conflicts of interest: how should a BCI company work alongside an AI company with links to the same founders?

Some governments have started looking at laws to protect “neural data”. This area is likely to get more attention as companies such as Merge Labs grow.

What Merge Labs Could Mean for Consumers

For now, don’t expect a Merge Labs headset in the Apple Store. Even the most advanced BCIs are still in medical trials, and consumer products face big technical, safety and regulatory hurdles. Still, the direction is clear. Tech companies want more natural ways to control devices.

You can already see that trend in today’s products. Apple Vision Pro uses your eyes and hand gestures instead of a mouse. Read our Apple Vision Pro overview for how that works. Apple has also added support for brain-computer interfaces in its accessibility features, so people using BCIs such as Synchron’s can control iPhone and iPad. See Apple Accessibility for the features available today.

If Merge Labs succeeds with a less invasive method, possible future uses could include:

  1. Helping people with disabilities communicate and use devices.
  2. Faster, hands-free control of computers and phones.
  3. New ways to work with AI assistants, beyond typing and voice.

Each of these depends on research results that haven’t been published yet. Watch for official announcements from the company rather than rumours.

Frequently Asked Questions

Who founded Merge Labs?

Sam Altman, the CEO of OpenAI, is a co-founder of Merge Labs, along with other co-founders named in news reports. Full details of the team and structure have not all been confirmed publicly.

Is Merge Labs part of OpenAI?

Reports describe Merge Labs as a separate company. It is linked to OpenAI through Altman, and some reports mentioned OpenAI’s interest as an investor, but it is not a division of OpenAI.

Does Merge Labs require brain surgery?

According to reports, Merge Labs is exploring less invasive methods, such as ultrasound, that may not need electrodes implanted in the brain. Its exact technology has not been confirmed publicly.

How is Merge Labs different from Neuralink?

Neuralink places electrode threads inside the brain through surgery and is already in human trials. Merge Labs is reported to be working on a less invasive approach and is at an earlier, less public stage.

When will Merge Labs release a product?

No product or release date has been confirmed. BCIs usually take many years of research and medical testing before they become widely available.

Conclusion

Merge Labs is one of the most interesting new companies in the brain-computer interface race. It pairs Sam Altman’s long-standing idea of humans and AI “merging” with a reported focus on less invasive technology. Many details are still unknown, and established players like Neuralink, Synchron and Precision Neuroscience are further along in human trials. Over the next few years, watch for real research results, clear rules on neural privacy, and medical use cases that help people first. Until then, today’s AI tools are the closest thing to working with AI, such as talking to ChatGPT hands-free on your iPhone.