In September 2021, Chile’s Senate approved a constitutional amendment by a vote of 37 to 0. The text was brief, and its subject was unusual for a constitution: the physical substrate of thought. Speech, property, and assembly have long-standing constitutional homes; brain activity did not. The amendment added a final clause to Article 19, Number 1 of the Chilean constitution, holding that scientific and technological development “shall be at the service of persons” and carried out “with respect for life and physical and psychic integrity,” and that the law must “especially safeguard cerebral activity, as well as the information derived from it” (Senado de Chile, 2021).
That a national constitution would single out brain activity forced a question that ethicists had pressed for years and that legislators are now being made to answer: if technology can infer what a person is thinking and, in some cases, influence how they feel, do the rights we already have protect them, or is something new required? The question is easy to state and hard to answer, because it mixes claims from three registers that public discussion constantly confuses. There is the demonstrated science—what laboratories have actually measured and replicated. There is plausible engineering—what could be built with known methods and enough resources. And there is speculation—the scenarios that motivate the debate but have no working path to them yet. A serious account of neural rights has to keep those registers separate, because the strongest argument for a new right looks quite different depending on which one you are standing in.
A constitutional amendment written by scientists
The Chilean reform followed sustained advocacy by scientists and legislators. One influential voice was the Morningside Group, a coalition of neuroscientists organized by Rafael Yuste, a Columbia University neurobiologist who had helped lead the United States’ BRAIN Initiative. The group had warned in Nature that brain-computer interfaces and AI could alter “core human characteristics: private mental life, individual agency and an understanding of individuals as entities bound by their bodies,” and it argued that existing human rights instruments were not built to address them (Yuste et al., Nature, 2017).
Chile was a plausible first mover. In 2023, its Supreme Court heard a case brought by the former senator Guido Girardi against Emotiv, a company whose consumer EEG headset required users to store their brain data in the company’s cloud. The court ordered Emotiv to delete Girardi’s neural data, cited the constitutional guarantee, and asked the national public-health institute to assess whether the device should be regulated as a medical product. The institute ultimately concluded that the Emotiv Insight did not fall under its medical mandate, because it delivers no diagnosis and no therapy (Biblioteca del Congreso Nacional de Chile, 2025). The episode captured the central legal difficulty in miniature: the most sensitive device in the case was not a medical instrument, and the general privacy regime was built for data that does not look like brain activity.
The amendment was only step one. A companion statute, Boletín 13.828-19, would define neural data, require prior, free, informed, and written consent before intervening in another person’s nervous system, mandate that neurotechnologies be “essentially reversible” unless used therapeutically, and classify neural data as a reserved and sensitive category. The June 2025 congressional briefing described the bill as being in its second constitutional stage. That dated record establishes the proposal and its procedural position then; it does not establish its present enactment status.
Four rights, and the claim that they are new
The intellectual foundation for these efforts was laid by two legal scholars. In 2017, Marcello Ienca and Roberto Andorno proposed four “neurorights” that they believed existing instruments did not adequately cover: the right to cognitive liberty, the right to mental privacy, the right to mental integrity, and the right to psychological continuity (Ienca & Andorno, Life Sciences, Society and Policy, 2017).
Their argument rested on a specific observation about where privacy law attaches. Conventional privacy protects information once it has been externalized—written, spoken, transmitted. Mental privacy, in their framing, would protect the information before it materializes, at the source. Mental integrity would guard against unauthorized intrusion into the mind, and psychological continuity would protect the sense of self from unconscious and unconsented alteration. Cognitive liberty—a term borrowed from the neuroethicist Wrye Sententia—would secure a person’s sovereignty over their own consciousness.
The distinction matters because it maps onto a real gap. A dataset of your search history is a record of what you did. A decoded pattern of brain activity may be a record of what you felt or intended, sometimes before you acted on it. If that inference is accurate, then the data does not merely describe you; it is closer to a readout of you. Whether that difference justifies a new right or only stronger handling rules under existing ones is the crux of the entire debate.
What the technology can actually do
The strongest case for neurorights is also where the field is most often oversold, so it is worth being precise about capability.
What has been demonstrated: invasive electrodes in clinical patients can decode intended speech and movement with useful accuracy, and can track mood-related signals well enough to guide deep-brain stimulation. Research teams have shown that structured mental content leaves a decodable neural trace under controlled conditions, with implanted hardware, and models trained per individual. This is real and often remarkable, and it is why the medical applications—restoring a voice to someone who has lost speech, reducing seizures, treating refractory depression—are genuine goods rather than marketing claims.
What is plausible engineering but not yet shown: consumer-grade, non-invasive devices that reliably infer specific thoughts, or that let a third party read a person’s emotional state from a headband or earbud with the confidence the word “read” implies. Non-invasive signals are attenuated and blurred by the skull, dosage is capped, and anatomy varies enough that targeting is imprecise. The gap between a laboratory decoder under ideal conditions and a wearable in a noisy room is large, and it is the gap that most policy hype ignores.
What remains speculation: undetectable, remote manipulation of a person’s beliefs or a device that inserts a manufactured experience without the person knowing. These scenarios drive legislation, and legislators are entitled to anticipate risk—but the honest accounting must say that the fear is currently ahead of the capability. The right response to that gap is not to dismiss the concern but to design rules that bite when the capability arrives rather than pretending it already has.
Surveys complicate the picture in a useful way. A nationwide U.S. study found that the public does not treat all brain data as equally sensitive: people rated some neural information as less private than a Social Security number but more private than media preferences, and they distinguished among the kinds of mental states being inferred (Huang et al., Journal of Law and the Biosciences, 2024). That finding cuts against a single blanket category and suggests that a workable regime may need to grade data by what it reveals, not merely by whether it came from a nervous system.
The commercial reality is already here
Whatever the science permits, the consumer market has already arrived. A 2024 report from the Neurorights Foundation surveyed thirty companies selling consumer neurotechnology and found that twenty-nine of the thirty appeared to have access to users’ neural data with no meaningful limitation; that only thirteen explicitly mentioned neural data in their policies; that more than 96 percent could transfer data to third parties; and that fewer than half extended users both the right to withdraw consent and the right to delete their data (Genser, Damianos & Yuste, 2024).
The report’s most important finding is not any single number but the ambiguity. Most companies have not decided whether neural data is a special category at all. Their privacy policies often describe a website rather than a device, leaving users uncertain whether the brain signals the product collects fall under any protection. This is the practical terrain on which the theoretical debate lands: not a world of mind-reading, but a world of unclear consent, weak deletion rights, and data that flows to third parties under ordinary commercial terms.
Consent is not a shield when the device sets the terms
The most persuasive argument for neural rights is not about exotic attacks. It is about a structural defect in how consent works when the intervention reaches the mind.
Ordinary consent law assumes a stable chooser: a person who can evaluate an option, agree or refuse, and remain the same person afterward. Neural technology can violate that assumption in two ways. First, the alternatives may be coercive—an employer, an insurer, a school, or a platform that conditions access on agreement. Second, and more troubling, the intervention can alter the very preference on which consent was based. If a device changes how a person feels about using the device, then “the user is satisfied” stops being a clean test of what the user would have chosen. Consent screens cannot repair that recursion.
This is why Chile’s draft statute pairs consent with reversibility, and why the UN Special Rapporteur on the right to privacy, Ana Brian Nougrères, has argued that neurodata is “highly sensitive personal data” requiring the precautionary principle, enhanced accountability, and “limited circulation” precisely because access to it can “negatively affect an individual’s mental integrity and thought processes” (OHCHR, 2025). Her report to the Human Rights Council made four recommendations to states: develop a specific regulatory framework for neurotechnology, incorporate right-to-privacy principles into national law, promote ethical practice, and educate the public so that consent can be genuinely informed.
The governance turn: from manifestos to instruments
The neurorights argument has moved, in under a decade, from an academic proposal to binding and near-binding instruments. Three are worth knowing because they define the current legal baseline.
In November 2025, UNESCO’s General Conference adopted the first global normative standard on the ethics of neurotechnology, at its 43rd session in Samarkand. The Recommendation treats the mind as inviolable in principle, calls on governments to keep neurotechnology inclusive and affordable, warns specifically against non-therapeutic use in children whose brains are still developing, cautions against workplace monitoring that builds productivity profiles of employees, and stresses the need to regulate products that may influence behavior or promote addiction (UNESCO, 2025). Its authority is soft-law, but its scope is genuinely global, and it gives member states a shared vocabulary.
In April 2025, the UN Human Rights Council adopted Resolution 58/6, “Neurotechnology and human rights,” requesting that its Advisory Committee develop guidelines for applying the existing human rights framework to neurotechnologies. The resolution’s language is telling: it anchors the response in established rights—including the absolute and unconditional character that human rights bodies have long attributed to freedom of thought—and requires consent that is prior, informed, free, explicit, transparent, effective, and revocable at any time. That is a different strategy from Chile’s: not new rights, but a more demanding reading of old ones.
At the national level, the United States has begun to move through a patchwork. Colorado became the first U.S. state to treat neural data as sensitive in 2024, followed by California, Montana, and Connecticut, each with subtly different definitions—some cover the peripheral nervous system, some only the central; some include algorithmically derived signals like heart-rate variability, some exclude them. In September 2025, a federal bill, the Management of Individuals’ Neural Data Act of 2025, was introduced by Senators Schumer, Cantwell, and Markey. It would not itself regulate; it would direct the Federal Trade Commission to study neural data governance and recommend a framework within a year (S. 2925, 2025). The bill is notable for the breadth of its definition—it reaches both central and peripheral nervous-system data, plus “other related data” like eye-tracking and voice analysis—and for its explicit aim of balancing innovation against abuse, including recommendations for use cases that might be prohibited “regardless of individual consent.”
The strong case against
It would be a mistake to present neurorights as settled. The most rigorous challenge comes from the legal scholar Jan Christoph Bublitz, who argues that the proposed rights should not be adopted and that lobbying for them should stop. His case is not that the technologies are harmless. It is that the proposed rights are poorly drafted, that calls for them risk “rights inflation” that devalues the currency of human rights, and that existing protections—bodily integrity, freedom of thought, privacy, the emerging doctrine of mental integrity—already cover the ground if properly interpreted (Bublitz, Neuroethics, 2022).
Related critiques sharpen the point. Some scholars argue that a right to mental integrity defined loosely enough to be useful becomes so broad that violations are everywhere, which trivializes the very harms it aims to address; others contend that rights-talk overemphasizes consent and thereby implies that reading or influencing a mind is acceptable whenever the person agrees—a confidence in consent that the coercion and recursion problems should undermine. A recurring worry is neuroexceptionalism: the assumption that brain data is metaphysically unlike every other kind of personal data, when the real harms often arise from combination—brain signals joined with location, purchase history, and behavior—rather than from the neural signal alone.
These are serious objections, and a fair account must concede their weight. The counterargument is practical rather than conceptual: even where a right is redundant in theory, a named and justiciable guarantee changes what a regulator, a court, and a company’s legal department will actually do. Chile’s amendment did not merely restate a principle; it gave its Supreme Court a hook it used, against a private company, within two years.
What would count as a right that works
If the debate is to produce something better than slogans in either direction, the useful test is whether a proposed protection changes behavior in a court or a compliance office. A few conditions recur across the serious proposals.
It must bind private actors, not only states. Most of the risk lives in commercial products, employment, insurance, and education, not in government surveillance. A right that constrains only the state leaves the dominant threat untouched.
It must be revocable and reversible. Consent that cannot be withdrawn, or an intervention that cannot be undone, is not a right but a one-way door. Chile’s reversibility requirement and the Special Rapporteur’s call for revocability at any time both point here.
It must reach inference, not just collection. The decisive harm is usually what can be concluded about a person, not the raw signal. Regulation that governs only the moment of acquisition will be bypassed by models that infer mental states from ordinary behavior.
It must distinguish degrees of sensitivity. The U.S. survey evidence and the divergence among state statutes suggest that a single category is both over- and under-inclusive. Some neural inferences may warrant near-absolute protection; others resemble ordinary sensitive data.
And it must preserve the benefits it exists alongside. Reading the field as a story of pure danger would be as distorting as reading it as pure promise. The same decoding that raises privacy alarms restores speech to people who have lost it and can make a treatment fire only when the patient’s own brain signals say it is needed. A regime that makes those applications harder to build, without making anyone safer, has failed its own test.
Where this leaves the mind
The scientific question and the legal question are moving on different clocks. The demonstrated capacities of neurotechnology are narrow, invasive, and clinical; the plausible engineering is broader but still bounded; the speculative scenarios that fill headlines remain speculative. Meanwhile, the governance apparatus is being built now—in a constitution in Santiago, a recommendation in Samarkand, a resolution in Geneva, and a patchwork of statutes from Denver to Sacramento. The instruments are converging on a common instinct even when they disagree about method: that a mind is not merely another dataset, and that the decision to read or shape one should not default to whoever holds the device.
Whether that instinct hardens into rights that individuals can actually invoke, or dissipates into declarations that describe a world no one is obliged to build, depends on details that are still being written: who is bound, what is revocable, what is inferred rather than measured, and whether the person whose mind is at stake retains the ability to refuse and to return to themselves. Those are not philosophical niceties. They are the difference between a right and a promise.
Sources and further reading
- Ienca, M. & Andorno, R., “Towards new human rights in the age of neuroscience and neurotechnology,” Life Sciences, Society and Policy, 2017
- Senado de Chile, “Histórica aprobación: información cerebral estará protegida en la Constitución,” 2021
- Biblioteca del Congreso Nacional de Chile, “Minuta: proyecto de ley sobre protección de los neuroderechos (Boletín 13.828-19),” 2025
- UNESCO, “Recommendation on the Ethics of Neurotechnology,” adopted 11 November 2025
- OHCHR, “UN expert calls for regulation of neurotechnologies to protect right to privacy,” 2025
- Genser, J., Damianos, S. & Yuste, R., “Safeguarding Brain Data: Assessing the Privacy Practices of Consumer Neurotechnology Companies,” Neurorights Foundation, 2024
- Huang, S. et al., “U.S. public perceptions of the sensitivity of brain data,” Journal of Law and the Biosciences, 2024
- Bublitz, J.C., “Novel Neurorights: From Nonsense to Substance,” Neuroethics, 2022
Loading comments…