UN Secretary-General recommends banning coercive use of neurotechnology in justice systems

By Inside BCI · October 11, 2026

Police in India have used a brain-wave test called Brain Electrical Oscillation Signature profiling, or BEOS, since the 2000s. A suspect wears an EEG headset while listening to statements about a crime, and software looks for signs that the person experienced the events described. In 2010 India’s Supreme Court ruled that forcing a person to undergo narcoanalysis, a polygraph or the Brain Electrical Activation Profile, a related brain-wave test, breaches the constitutional protection against self-incrimination, and that the results cannot on their own be admitted as evidence. Testing with consent, recorded before a magistrate, remains allowed, and some suspects ask for BEOS to try to clear their names.

UN Secretary-General António Guterres now recommends that states prohibit the compulsory or coercive use of any neurotechnology in justice systems, and hold any claimed consent to the highest evidentiary standard. His recommendations, in a report to the General Assembly’s eighty-first session, do not bind governments and carry no timetable. They also reach beyond police equipment: he asks states to treat neural data as highly sensitive personal data, limit its sale and reuse, and press neurotechnology companies to carry out human rights due diligence.

What Guterres recommends

The report, Human rights in the administration of justice (A/81/351), is dated 27 August. The General Assembly asked for it in a resolution adopted on 17 December 2024, which sought the latest developments, risks and required safeguards on the potential use of neurotechnology and other emerging technologies in the justice system. The previous report in the series, in 2024, gave neurotechnology one short section in a wider review of digital technology and artificial intelligence. This one is devoted to it.

Its conclusions open with the line that neurotechnology “raises profound human rights challenges when applied to the administration of justice”, and that its spread from clinics into investigative, judicial and correctional settings brings significant risks alongside potential benefits. Guterres makes ten recommendations to member states. Five deal with coercion, abuse, consent, evidence and data.

States should prohibit the compulsory or coercive use of neurotechnology in the administration of justice, including brain-based deception detection, predictive applications, decoding applications that claim to read thoughts, intentions or memories, neuromodulation, and arbitrary or unlawful cognitive surveillance and monitoring.

States should prohibit the use of neurotechnology as a means of torture or ill-treatment, or for any kind of social control or coercive behavioural conformity, and bar neural data obtained through torture or ill-treatment, or in violation of the rights to freedom of thought and opinion, from use as evidence in any proceedings.

Any use of neurotechnology in justice settings should require free, prior, informed, explicit and revocable consent, with full disclosure of risks, access to an independent lawyer of the person’s own choosing and routes to redress. Because justice settings are constrained and coercive, any such consent should face the highest evidentiary standards and rigorous scrutiny.

Neurotechnology-derived evidence proposed for admission in court should undergo rigorous independent scientific review and validation, including, as appropriate, of its safety, validity, reliability, reproducibility, known limitations and suitability for the context and population concerned. Any lawful deployment should come with oversight, monitoring and public reporting.

States should recognise neural data as highly sensitive personal data with the highest level of legal protection, and set clear limits on its collection, processing, sale, storage, retention, reuse and sharing, backed by remedies and independent oversight.

The other five ask states to set up regulatory frameworks for neurotechnology used in justice systems, covering its development by companies as well as its use by the state, with human rights impact assessments and due diligence across its life cycle and testing for algorithmic bias; to add safeguards for children, detainees, people with disabilities and other groups; to respect the rights of Indigenous Peoples; to urge neurotechnology companies to meet their human rights responsibilities; and to invest in public education and in training for police, prosecutors, judges and defence lawyers.

Where neurotechnology already reaches the courtroom

The most common use today is brain imaging offered as medical or expert evidence. In criminal cases it mostly supports arguments for a lighter sentence or claims about a defendant’s mental capacity. In civil cases it is used to assess non-financial damages, to support claims of chronic pain or cognitive impairment, and to examine legal capacity. Guterres cites practice in Brazil, the Netherlands, the United States, Italy, England and Wales, and Canada, and finds that where courts accept such evidence they generally treat it as corroboration for other evidence.

Brain-based lie detection is rarer. No method has gained general acceptance for courtroom use, and in 2012 a US federal appeals court upheld the exclusion of functional MRI lie-detection evidence as insufficiently reliable. Guterres adds that neurotechnology cannot yet read minds, though it is reportedly able to infer emotions and internal reactions from neural data.

Predicting reoffending from neural markers and decoding thoughts, intentions or memories during interrogation are described as experimental or likely to be explored, and as raising serious concerns for the presumption of innocence, privacy and freedom of thought. Wearable monitoring of offenders outside custody and brain stimulation to curb aggression or impulsivity are listed as anticipated developments whose human rights issues should be addressed before any deployment.

What it means for consumer neurotechnology companies

Wearable EEG devices sold for wellness, gaming and productivity are widely available and face little or no regulation outside the medical field. Guterres warns, citing outside research, that the neural data they collect may be accessed by state authorities or introduced as evidence without regard to its sensitivity, a risk he says is greater in counter-terrorism and national security cases, where legal thresholds for data access and admissibility are often lower.

The report draws on an April 2024 Neurorights Foundation review of 30 consumer neurotechnology companies. That review found that 29 appeared to have access to users’ neural data with no meaningful limits on that access, that 17 said in their policies they could share consumer data to comply with legal requests, and that none said it would resist requests from law enforcement or government.

The recommendation on business is addressed to governments, which Guterres asks to urge every company involved in neurotechnology to meet its responsibility to respect human rights under the UN Guiding Principles on Business and Human Rights, through human rights due diligence, transparency, plain-language consumer information, safety by design and access to remedy.

Where this sits on the four-jurisdiction posture map

Issue 01 of the Dargentic Intelligence Report classified four global regulatory postures on brain-computer interfaces: Rights-First (Chile), Horizontal Regulator (EU with AI Act and GDPR), State-Patchwork (US), and Industrial Builder (China and South Korea). Nine governments’ submissions are published with the call for input behind the report: Argentina, Austria, Brazil, the Dominican Republic, Italy, Morocco, Slovenia, Ukraine and Uzbekistan. The United States, China and South Korea are not among them, and the survey of national law that results touches the four postures unevenly.

Rights-First. Guterres highlights Chile’s 2021 constitutional amendment protecting brain activity and the data derived from it, which the Supreme Court applied in August 2023 in a case against Emotiv over a portable brain-monitoring headset. Outside Chile, he notes a proposed general law on neurorights in Mexico.

Horizontal Regulator. The EU AI Act can classify AI-driven neurotechnology and emotion recognition as high-risk. AI systems used by police as polygraphs or similar tools are high-risk under the Act where EU or national law permits their use, with obligations that apply from 2 December 2027 after this year’s deferral. Since February 2025 the Act has also prohibited AI systems that predict a person’s risk of offending based solely on profiling or personality traits. Guterres’s recommendation would add a prohibition on any compulsory or coercive use of neurotechnology in justice systems, with or without artificial intelligence.

State-Patchwork. The United States has no federal law specific to neurotechnology, and the report names California, Colorado and Montana as states that have extended sensitive data protections to neural data. It omits Connecticut, whose law took effect on 1 July, and Vermont, which in June made neural data sensitive data requiring consent to process or sell from 1 January 2028. Delaware added neural data to its sensitive data category on 2 September, after the report’s date, with effect from 1 January 2027. On 27 September California’s governor vetoed AB 1542, which would have banned the sale of sensitive personal information, so the sale of neural data in California remains subject to a consumer opt-out. Three days later he signed AB 1883, which from 1 January 2027 bars employers from using AI workplace tools to collect workers’ neural data.

Industrial Builder. The survey’s only reference to China is a set of 2024 ethical guidelines for brain-computer interface research covering research integrity, scientific validity and informed consent. It does not cover South Korea.

What happens next

The General Assembly decided in 2024 to return to human rights in the administration of justice at its current session. Separately, the Human Rights Council has asked its Advisory Committee to draft guidelines on applying existing human rights standards across the neurotechnology life cycle, for the Council’s sixty-fourth session in early 2027. The UN Office on Drugs and Crime and Interpol began joint work in 2025 to explore potential uses of neurotechnology in law enforcement and criminal justice, and the associated risks.

Not disclosed

The report does not estimate how many police forces, courts or prison systems use neurotechnology. It names no companies selling brain-based deception tests, though it identifies two commercial labels, brain fingerprinting and BEOS. It cites no case in which police obtained neural data from a consumer device or a court admitted such data as evidence.

What to watch

Whether the General Assembly’s next resolution on human rights in the administration of justice takes up the recommendation to prohibit compulsory or coercive use, and the neural data language.

Whether the Human Rights Council Advisory Committee’s draft guidelines carry the same prohibitions.

Whether the UN Office on Drugs and Crime and Interpol publish guidance for police on neurotechnology, and what consent standard it sets.

Whether consumer neurotechnology companies publish how they respond to law enforcement requests for neural data.

Sources

Primary:

Coverage:

Cross-reference to prior InsideBCI coverage:

policy United-Nations Antonio-Guterres UN-General-Assembly A-81-351 OHCHR human-rights administration-of-justice neural-data brain-based-lie-detection BEOS neuromodulation freedom-of-thought consumer-neurotechnology Neurorights-Foundation EU-AI-Act Chile-neurorights UNODC Interpol four-posture-map Issue-01