After reading about, experiencing and analysing the current technical and regulatory landscape of brain-computer interfaces (BCIs), and speaking with technologists, policy makers, legal scholars, and data experts, we encapsulated our learnings into eight recommendations for how to shape the ‘tech we want’ for our brain commons.
Share your feedback with us here.
1. Understand the Neurotech Stack
Millions of people will acquire commercial neurotech devices in the coming years. As the global market for these technologies expands, our first recommendation is to dedicate effort to understanding what brain data is and how consumer technologies work. The gap between the marketing claims and what the products can actually do is significant. A widespread misconception is that BCIs can read minds. In reality, consumer headsets that measure the electrical activity of the brain cannot detect the content of thoughts, memories, or imaginations. What they can try to infer is levels of stress, attention, or various emotions—for instance, in combination with measurements of heart rate, eye tracking, and behavioral data from other sources. Technical knowledge gaps are ripe for exploitation by industry. Companies will claim to deliver deep cognitive understanding while actually only performing simple pattern matching. The devices may be sleek and expensive, but we should be skeptical of their sci-fi-inspired marketing.
2. Follow the Money
We need a better understanding of who is investing in neurotech and what their long-term plans are. Major investors and tech giants have already sketched out future plans for the development and sale of neurotech products—and they have the power to hype them, regardless of what they do (or should do) in reality. The consumer market is growing rapidly with limited regulatory frameworks. Investment trajectories for tech are usually locked in place before public interest voices ever enter the conversation. This is happening now too. Among the many risks, legitimate medical breakthroughs could be kept out of reach of global majority populations. Brain research could become even more militarised, and policies will be driven by industry interests rather than public needs.
3. Resist the Temptation to Create Specialised Regulation
We should anchor BCI governance in established public interest technology frameworks and public law rather than through dedicated neurotech legislation. Today, BCIs products are emulating cloud and mobile ecosystems by creating anti-competitive vendor lock-ins. In the past, competition law, consumer protection law, children’s protection frameworks, and robust interoperability standards have often proven more effective for preventing harms than dedicated legislation that can be manipulated by industry lobbying. Concerns associated with BCIs—including biased inferences and alteration of mental states—are shared across other technologies too. Treating tech and data governance problems as being unique to a specific type of technology will lead to policy gaps that industry will exploit. Regulation anchored to new concepts of “neural data” would miss the broader class of risks. Tech-neutral framings that focus on the intent to infer mental states—regardless of the source of data—will be harder to circumvent.
4. Abolish all Notions of a “Normal” Brain
We should be wary of any BCI architecture that uses “normal” brain function or activity as a departing point of metrics for measurements, interpretations or inferences of cognition. The uniformity pushed by the neurotech industry is an erasure of human dignity and cognitive diversity. People should not have to adapt to corporate-defined models of cognition. This is epistemically violent, regardless of whether it may be commercially practical. A public-interest BCI ecosystem must treat diversity in brain function as a baseline requirement.
5. Hold BCI Developers to Account
BCIs have lots of potential to improve people’s lives in different ways. However, we have seen in the past how harmful technologies are at first presented as medical or accessibility aids to help people in need, before being deployed on a mass scale for a different purpose. If a BCI makes genuine claims to serve a disability community or perform an assistive function, it must also be held to strict design standards: the simplest, lowest-energy, most portable, scalable, and maintainable solution available—especially when public funds are involved in financing the research. We need to question every product that claims to benefit people who are vulnerable, and we need to ensure that BCI manufacturers are held responsible for any harm, regardless of intent. Standard open-source licenses, which often disclaim liability are insufficient. Regulators should apply “objective liability” to all BCIs, not just medical devices.
6. Demand Openness by Default
Openness is a practical governance mechanism for any technology, at every layer. It is especially relevant for BCIs. Focusing on privacy alone is not enough. We need to discuss standards, procurement and platform power too. Beyond code, it refers to the auditability, portability, and user control of technologies. Without it, we cannot fully know the real capabilities (or double-use cases) for neurotech. Everyone should have the right to transfer their trained neural profiles if they switch hardware, and should have the right to know how their devices are made so they can repair them or connect them with others. We need to counter opacity by design—enabled by trade secrets, closed firmware, proprietary data formats, and cloud-only functionality. We can do it by embedding openness into standards from the outset, but also through creative circumvention of the current patent system, for instance, through new exceptions to trade secret law or with specific administrative regulation for devices.
7. Protect the BrainCommons
Patient community practices should be strengthened to ensure that neural infrastructure can serve a true public interest purpose. By developing more pathways for communities to govern neural data, we can ensure that critical neural infrastructure becomes a shared resource. Protection of a BrainCommons could involve test patent pools or data trusts owned by the affected communities, who would also be involved in processes to design, test and fine tune technologies. Collective neural data should not be locked behind proprietary walls or hidden in the noise of commercial platforms.
8. Protect Freedom of Thought
We should not wait for harm at scale to regulate. The right to “freedom of thought” under Article 18 of the Universal Declaration of Human Rights is absolute. It protects the right to keep thoughts private, and the right not to have thoughts manipulated. BCIs that infer or alter mental states operate directly within this protected space. Those seeking to deploy these technologies should have to justify their use. We should push for preemptive prohibitions of specific BCI applications, especially when there is no meaningful opportunity to opt out: prison control, workplace cognitive monitoring, military performance optimization, student attention profiling, and any other coerced adoption. We need to act now before such uses are normalised further as cost efficient or necessary. People must have a meaningful right to opt out of neural monitoring.