OpenAI Invests 252 Million Dollars in BCI Startup Merge Labs

Human fingers are too slow to keep pace with artificial intelligence. Even voice commands act as a bottleneck compared to the speed at which the brain thinks. Now, OpenAI seeks to overcome this limitation not through software updates, but by directly connecting biological neural networks with silicon chips.

OpenAI, led by Sam Altman, has made a $252 million (approximately 330 billion KRW) seed investment in the brain-computer interface (BCI) startup 'Merge Labs.' This decision reveals OpenAI's ambition to expand its ecosystem beyond agents that generate text and images, toward the hardware-based extension of human biological intelligence itself. The AI fever that began with ChatGPT is now targeting the final frontier: the human brain.

Merge Labs' Gamble: Choosing Ultrasound Over Needles

The technology proposed by Merge Labs differs from Elon Musk’s Neuralink. While Neuralink adheres to an invasive method of implanting electrodes directly into the brain, Merge Labs is developing a non-invasive interface that combines ultrasound and molecular engineering. The core of this technology lies in 'gene therapy.'

Merge Labs uses a method of creating genetic 'gas vesicles' that cause specific neurons to respond to ultrasound. When a wearable ultrasound device worn on the head detects or modulates these vesicles, it can read brain signals or deliver specific stimuli. The high-bandwidth neural data collected this way is decoded in real-time by OpenAI's artificial intelligence models.

Currently, this technology remains in the early research and development (R&D) stage. As of January 2026, Merge Labs has just emerged from stealth mode at the seed stage, and actual commercialization is expected to take at least 10 to 15 years. This suggests that the investment is based on a long-term vision of expanding human cognitive capacity rather than immediate profit.

An Operating System for the Brain: 'Scientific Foundation Models'

OpenAI's interest in Merge Labs is not merely due to hardware. The two companies plan to co-develop an 'AI Operating System (AI OS)' and 'Scientific Foundation Models' that precisely interpret the intent of brain signals and adapt to individual characteristics.

Biological signals are noisy and fragmented. OpenAI’s models will be tasked with identifying the user’s intentions and emotional states within this raw data. If successful, humans could communicate with AI solely through thought, without physical input devices like keyboards or mice. This signifies an era where AI becomes an extension of intelligence rather than just a human tool. The strategy is to fundamentally enhance human cognitive abilities by accelerating bio-engineering and neuroscience research.

The Shadow of Cognitive Enhancement: Vanishing Privacy and Agency

Behind this rosy outlook lies a massive ethical pitfall. The greatest concern is 'neuro-privacy.' Data collected through BCI includes not only conscious thoughts but also unconscious emotions and desires. When such data is accumulated on corporate servers and used commercially, the human inner world can no longer remain a secret.

The ambiguity of behavioral agency is also an issue. If AI is directly connected to the brain to assist or intervene in decision-making, can the result still be called a purely 'human decision'? The boundaries of the self risk collapsing as AI recommendations and human will intertwine. Furthermore, intellectual inequality between those who can afford cognitive enhancement technology and those who cannot is likely to become a new source of social conflict.

What We Need to Prepare For

While Merge Labs' technology is still in the lab, developers and researchers must already prepare for this new interface. Future software will evolve from designing visual information displayed on screens to designing neural stimuli delivered directly to the brain.

Currently, clinical trial schedules and approval from regulatory authorities remain unclear. However, given the combination of OpenAI's massive capital and technical prowess, discussions on encryption standards for neural data and ethical guidelines must be prioritized.

FAQ: Frequently Asked Questions

Q1: Is the ultrasound method safer than implanting electrodes directly into the brain? A1: Theoretically, yes. Since it is a non-invasive method that does not require opening the skull, the risk of infection or brain tissue damage is significantly lower. However, long-term safety verification regarding the process of permanently modifying brain cell properties through gene therapy is essential.

Q2: Will my thoughts be stored on OpenAI servers in real-time? A2: OpenAI and Merge Labs have not yet released specific details regarding privacy and data security policies. Given the sensitivity of neural data, 'on-device' processing and robust encryption standards will likely need to be implemented.

Q3: When is the commercialization expected to occur? A3: The technology is currently in the early R&D stage. Research teams and market experts anticipate it will take another 10 to 15 years for consumer-grade devices to reach the general public.

Conclusion: The Era Where the Boundary Between Silicon and Neurons Dissolves

OpenAI's investment in Merge Labs is not a simple business expansion. It is a bold attempt to incorporate artificial intelligence as part of the internal neural network rather than an external tool. We are moving toward an era where the definition of 'intelligence' is no longer confined by biological limits. This presents an opportunity for a leap in human cognitive capacity, while simultaneously posing a grave challenge to redefine 'what it means to be human.' We must closely monitor how OpenAI controls the power of this potent technology and establishes ethical guardrails.

Aionda