Neurorights and the Ethics of Reading the Brain

Rafael Yuste (Columbia, NeuroRights Foundation) and colleagues have proposed a five-right framework that has become the de facto policy reference worldwide: (1) personal identity, (2) agency, (3) mental privacy, (4) fair access to mental augmentation, and (5) protection from algorithmic bias.

Chile became the first country to constitutionally protect neurorights through a 2021 amendment to Article 19 of its Constitution. Spain has incorporated digital and neurorights into its 2018 Digital Rights Charter; Brazil, Mexico, and Argentina have all advanced parallel legislation through 2023-2025.

UNESCO adopted a global Recommendation on the Ethics of Neurotechnology in November 2025 - the first instrument of its kind at the UN level. The OECD's 2019 Recommendation on Responsible Innovation in Neurotechnology remains the leading framework for member-state policy.

Even crude consumer EEG (Muse, Emotiv, Neurable, Mendi) can reveal information about emotional state, attention level, certain medical conditions (epileptiform activity, sleep architecture), and reactions to specific stimuli. The technical possibility of involuntary inference about mental content - what a person is attending to, whether they recognise a face, whether they intend to deceive - is no longer purely theoretical for some narrow tasks.

Colorado HB 24-1058 (signed April 2024) became the first US state law to explicitly classify neural data as 'sensitive personal information' under that state's privacy framework. California's Senate Bill 1223 (signed September 2024) added neural data to the categories protected under the CCPA. Minnesota and New Hampshire introduced similar bills in 2025.

The EU's GDPR already treats brain-related health information as a special category of personal data. The proposed EU AI Act's high-risk categorisation and Article 5 prohibitions further constrain emotion recognition and biometric inference in some contexts.

Closed-loop stimulation that modulates mood or behaviour raises identity-altering possibilities and complicates legal notions of responsibility, consent, and authenticity of action. The literature on DBS-induced personality change is small but growing; a minority of DBS recipients report altered self-perception or interpersonal dynamics post-implant.

These reports are reshaping informed-consent practice. Many DBS programs now use formal preoperative neuropsychiatric assessment, post-operative monitoring for personality change, and structured shared-decision-making tools.

Legal scholarship (Bublitz, Merkel; Ienca; Wexler) explores how concepts such as cognitive liberty, mental integrity, and authenticity of decision-making should translate into criminal-responsibility and contract-law doctrines as neurotechnology becomes more capable.

Implanted BCIs in current clinical practice cost USD 100,000 to 500,000 per patient. Access depends on insurance, geography, surgical infrastructure, and clinical-trial enrolment - a combination that risks a sharp 'neuro-divide' if and when therapies become elective enhancements.

The WHO Global Initiative on Brain Health, UNESCO's neurotechnology workstream, and the OECD's Working Party on Bio- Nano- and Converging Technologies are coordinating to set minimum norms before consumer neurotech proliferates.

Low- and middle-income country access to even mature neuromodulation therapies (cochlear implants, DBS) remains highly uneven; the WHO estimates that the global cochlear-implant gap exceeds 90% of eligible candidates in many regions.

Employer-mandated EEG monitoring is no longer hypothetical. Documented deployments in long-haul trucking, mining, and manufacturing - particularly in Australia and China - have raised consent and surveillance concerns now, not in the future.

Italy's data-protection authority (Garante) and Chile's neurorights regulator have issued enforcement actions against specific consumer-neurotechnology deployments. The European Data Protection Board is preparing guidance on neural data under the GDPR.

The Future of Privacy Forum's 2024 brain-data report and the Neuroethics Working Group's recommendations are emerging as practical reference standards for employers, schools, and consumer-product companies.

Wireless implanted devices have been shown in security research to be vulnerable to passive eavesdropping and, in some demonstrated cases, active interference (Halperin et al., IEEE 2008, on cardiac implants; subsequent work on insulin pumps and DBS systems). Industry response has improved encryption and authentication, but the surface area continues to grow.

Adversarial machine learning attacks against neural decoders are a distinct and underexplored threat class. Robust decoding under adversarial conditions is an active research direction.