Neural Data Is Not Ordinary Data

My illustration entitled: “Memories Under Glass” – Personal memories appear as fragile, glowing scenes inside transparent neural capsules, watched by distant surveillance systems.

Why Information from the Nervous System Requires Exceptional Protection

Digital systems already collect vast quantities of information about human beings. Search histories reveal curiosity. Location records reveal movement. Purchase histories reveal consumption. Messages reveal relationships, while metadata can expose the structure of a person’s social life without disclosing the contents of a single conversation.

Neural data introduces a different category of information.

It originates from the activity of the brain or nervous system. Depending upon the device and context, it may be used to identify patterns associated with movement, attention, recognition, cognitive workload, emotional response, health or intention. Artificial intelligence can convert these complex signals into classifications, predictions and commands that may become more revealing as analytical techniques improve.

A browser history records an action a person has already performed. A neural interface may detect activity involved in preparing an action before that action is externally expressed. A camera records a visible reaction. A neural sensor may attempt to estimate processes that remain internal. A questionnaire asks a person to provide an answer. A brain–computer interface may generate information without requiring the person to formulate a deliberate statement at all.

This does not mean that contemporary neurotechnology can transparently read every thought. Neural signals are complex, noisy and dependent upon interpretation. Exaggerated claims about mind reading can distort public understanding and obscure the real capabilities and limitations of the technology.

Nevertheless, uncertainty does not make neural information ordinary. An incomplete or inaccurate cognitive inference can still be used to classify, reward, exclude, persuade or discriminate against a person.

The distinctive importance of neural data lies not in science-fiction claims of perfect mind reading, but in its origin, intimacy, involuntary production, inferential potential and connection to technologies capable of acting upon the nervous system.

Neural data is not ordinary data because the nervous system is not an ordinary data source.

What Is Neural Data?

Neural data is information generated through the measurement of activity, structure or function within the central or peripheral nervous system that can be recorded or processed by a device.

This broad category may include signals collected through electroencephalography, implanted electrodes, neural prostheses, brain-imaging systems, peripheral nerve interfaces and other neurotechnologies. It may also include information derived from those signals after processing by software or artificial intelligence.

For the purposes of Neuro-Cypherpunkism, the category should be divided into several layers:

Raw neural data: signals recorded directly from the brain or nervous system before substantial interpretation.
Processed neural data: signals that have been filtered, compressed, segmented or converted into features usable by software.
Inferred cognitive data: conclusions concerning intention, attention, recognition, emotion, fatigue, preference or other mental states derived from neural information.
Neural command data: processed outputs used to control a cursor, prosthetic device, application or other connected system.
Neural write data: instructions or parameters governing stimulation directed toward the nervous system.
Neural metadata: information about when, where, how and through which device neural activity was recorded or transmitted.
Learned neural profiles: models developed over time to recognize the distinctive signal patterns of a particular individual.

These layers should not be treated as interchangeable. A game may need to receive a simple command without receiving the raw neurological recording that produced it. A clinician may require access to particular signals for treatment without requiring permanent commercial rights over the patient’s learned neural profile.

Separating these categories allows access to be limited according to purpose. It also prevents consent to one useful function from becoming permission to extract everything the device can detect.

1. Neural Data Originates from the Person

Ordinary consumer data is frequently generated through an external transaction. A person purchases an item, visits a website, enters a location or sends a message. Neural data originates from biological activity within the person.

A manufacturer may own the physical sensor it produces and retain intellectual-property rights over its software. Those rights do not give it moral ownership of the nervous-system activity detected by the device.

The distinction resembles the difference between owning a camera and owning everything the camera observes. Providing the instrument does not establish ownership over the subject. With neural technology, the case for individual authority is even stronger because the information originates from processes closely connected to cognition, bodily function and identity.

The collection of information from the person must not become ownership of the person.

2. Neural Data Can Be Produced Without Deliberate Expression

Many conventional forms of personal information involve deliberate expression. A person types a sentence, completes a form, posts an image or selects a product. Neural information may be generated without a conscious decision to communicate.

A brain–computer interface may continuously receive signals while the user operates it. Some of those signals may be relevant to the requested function; others may be incidental. The individual may not know which patterns are technically available or what future analytical methods could derive from them.

This weakens the assumption that collection equals disclosure. A signal emitted by the nervous system is not necessarily a statement intended for an audience. The fact that a sensor can detect it does not mean the person has chosen to reveal its meaning.

Consent to wear or implant a device must therefore not be interpreted as unlimited consent to analyse every signal that becomes accessible.

3. Neural Data Is Inference-Rich

The importance of neural information often lies not in the raw recording but in what an algorithm claims to infer from it.

A neural signal may initially appear to be an unintelligible sequence of measurements. Machine-learning systems can identify patterns across time, compare them with training data and generate outputs concerning motor intention, recognition, attention or other characteristics.

This creates a distinction between disclosed information and inferred information. The individual may authorize a device to perform one operation while the system derives additional conclusions that were never knowingly volunteered.

The problem resembles the transformation described in Artificial Intelligence and the End of Informational Privacy. AI can combine fragments of information to produce knowledge the individual never directly disclosed. With neural data, those fragments originate closer to the biological processes underlying perception and action.

Privacy protection must consequently extend to derived information. Deleting raw data while retaining a detailed cognitive profile does not restore privacy. An inference about a person can be as consequential as the recording from which it was generated.

4. Neural Data May Precede External Action

Most behavioural data records what a person has already done. Neural technology may detect signals involved in the preparation of movement, speech or response before an externally observable action occurs.

This does not establish that every neural signal represents a settled decision. Human cognition contains hesitation, imagination, discarded possibilities and involuntary reactions. A preliminary neurological pattern may never become an action or belief.

That is precisely why neural information requires caution. A person should not be judged as though every detected pattern were a completed intention. Institutions must not confuse neurological possibility with deliberate choice.

Law and morality ordinarily distinguish thought from conduct. A technological system should not erase that distinction by treating pre-action neural activity as if it were a confession, commitment or completed act.

5. Neural Data Is Context-Dependent and Easy to Misinterpret

Neural signals do not carry a fixed meaning independent of context. Their interpretation may depend upon the individual, device, environment, task, calibration process and statistical model.

A pattern classified as fatigue in one setting may have a different explanation in another. A signal associated with recognition does not necessarily establish familiarity, agreement or intent. Emotional and cognitive categories imposed by software may simplify states that are complex, overlapping and culturally influenced.

This makes neural classification dangerous when used outside the context in which the system was developed. A model designed for clinical rehabilitation should not automatically be assumed reliable for employment screening, criminal investigation or insurance assessment.

Neural-data governance must therefore address accuracy, uncertainty and purpose. Systems should disclose their limitations, and consequential classifications must remain open to challenge.

An algorithmic interpretation of neural activity is an inference produced by a system. It is not an infallible statement of who the person is.

6. Neural Data Can Become Persistent

A password can be changed after a breach. A payment card can be cancelled. Some neural characteristics may be more difficult to replace because they arise from the person’s own biology and learned interaction with a device.

A brain–computer interface may be calibrated to recognize an individual’s signal patterns. Over time, the system may accumulate a model that becomes increasingly valuable for authentication, prediction or control. If this learned profile is compromised, the individual may not be able to generate an entirely new neurological identity.

Persistence also operates through storage. Neural information collected today may be reanalysed by more capable algorithms in the future. A dataset believed to reveal little at the time of collection may later support conclusions that neither the user nor the original researcher anticipated.

Consent must therefore account for duration and future analysis. Permission for a present function cannot become unlimited permission for every future capability.

Data retention should be limited according to necessity. Where continued storage is justified for clinical or scientific purposes, access should remain secure, auditable and governed by clearly defined conditions.

7. Neural Data Can Affect Bodily and Cognitive Integrity

Ordinary personal data is generally treated as information flowing away from the individual. Neural systems may form part of a two-way channel.

A brain–computer interface may read activity from the nervous system, while a connected neurostimulation system may transmit influence toward it. This creates a direct relationship between information security and bodily integrity.

An attacker who steals a shopping history violates privacy. An attacker who alters a command governing neural stimulation may create physical or psychological harm. Even when read and write functions exist in separate components, the ecosystem connecting them must be secured as a whole.

This is why The Twelve Principles of Neuro-Cypherpunkism establish a strict neural read/write divide. Authorization to receive information from the brain must never imply authorization to send stimulation toward it.

Neural cybersecurity is not merely data protection. It can become protection of the person.

8. Neural Data Can Create Exceptional Dependency

Some people will use brain–computer interfaces as optional consumer devices. Others may depend upon them to communicate, control a prosthetic limb or interact with their surroundings.

This dependency changes the relationship between the individual and the technology provider. A person may be unable to reject new terms, leave the platform or stop transmitting information if doing so also removes an essential human capability.

A manufacturer could theoretically control software updates, cloud access, compatible applications and continued device functionality. If the company ceases operating or withdraws support, the user may lose more than an ordinary service. He may lose a means of speech, mobility or independence.

Neural-data protection must therefore include continuity, portability and the right to change providers. A user should not become a permanent tenant inside a technological system connected to his own nervous system.

Where dependency is greatest, institutional responsibility must be strongest.

9. Neural Data Can Be Used to Exercise Power

Information becomes politically important when it changes what institutions can do to individuals.

An employer may use inferred attention to evaluate workers. An insurer may attempt to interpret neural indicators of health or risk. An advertiser may seek to measure neurological responses rather than relying on clicks. A government may view neural information as evidence of recognition, intention or deception.

Some applications may be presented as efficient or objective. Yet the apparent precision of neurological measurement can conceal uncertainty and institutional bias. A technologically produced score may acquire authority far beyond what its scientific reliability warrants.

The central concern is not only exposure. It is asymmetry. The institution possesses the device, data, model and decision-making authority, while the individual may not know what was inferred or how to challenge it.

As argued in Metadata Is Power, information creates leverage when one party can observe and classify another without equivalent visibility or control. Neural information may intensify that asymmetry by making private biological activity available to external institutions.

10. Neural Data Can Be Combined with Everything Else

The risks of neural data cannot be evaluated in isolation. A neural signal may reveal little on its own but become highly informative when combined with identity, location, medical history, video, speech, purchasing behaviour and social relationships.

Artificial intelligence can connect these categories and build models more revealing than any individual dataset. Neural information may confirm, refine or contradict conclusions derived from observable behaviour.

Anonymization may offer protection, but it is not a universal solution. Distinctive patterns and external datasets can sometimes enable information to be linked back to an individual. A learned neural profile designed for personalization may itself function as an identifier.

Data protection should therefore evaluate the entire information environment. A company should not claim that neural data is harmless because names have been removed while retaining the signals and associated records necessary to reconstruct identity.

11. Neural Data Can Concern More Than One Person

Neural information originates from an individual, but some conclusions derived from it may also affect families, communities or groups.

Research may identify patterns associated with a medical condition shared by biological relatives. Models trained on one population may be used to classify others. Group-level predictions may influence how institutions treat people who never contributed data to the original system.

This does not eliminate individual authority. It demonstrates that consent and privacy cannot be treated only as transactions between one consumer and one company. Neural-data systems may produce collective consequences requiring scientific oversight, anti-discrimination protections and public deliberation.

The individual remains the primary subject of cognitive sovereignty, but responsible governance must also consider people indirectly affected by the resulting models.

12. Neural Data Is Connected to Human Identity

Not every neural measurement reveals a person’s deepest self. Some signals may perform narrow and ordinary functions. A simple command used to move a cursor may reveal less than a detailed search history.

Neural exceptionalism should therefore not become scientific mysticism. Brain-generated data is not sacred merely because it originates in the nervous system, and different forms of neural information present different levels of risk.

Yet the category deserves special protection because it is connected to the biological processes underlying perception, intention, memory, emotion and action. As devices become more capable, their data may become increasingly relevant to how institutions describe and judge the person.

A neural profile can become part of a technological representation of identity. If that profile determines access, predicts behaviour or mediates communication, control over it becomes control over how the individual exists within the system.

The person must not lose authority over his technological identity simply because an algorithm helped construct it.

Not All Neural Data Carries the Same Risk

Recognizing neural data as exceptional does not mean that every measurement should receive identical restrictions. A risk-based framework is necessary.

A transient signal processed locally to move a character in a game may create limited privacy risk if the raw data is immediately discarded. A long-term clinical recording connected to identity and medical history requires stronger safeguards. A neural profile used to infer emotion for employment decisions creates a different and potentially more serious danger.

The level of protection should consider:

the intimacy of what the data can reveal;
whether collection is continuous or occasional;
whether the information was deliberately expressed;
whether raw signals or only limited outputs are retained;
whether the data can identify or profile the individual;
whether artificial intelligence generates additional inferences;
whether the system contains neural write or stimulation capabilities;
whether the individual depends upon the device;
whether the information is used in medicine, employment, insurance, education or law enforcement;
whether errors can produce significant harm; and
whether the person can withdraw consent or leave the system.

A proportionate framework protects individuals without treating every benign experiment as an intolerable intrusion. The objective is responsible innovation, not prohibition.

Why Existing Consumer Consent Is Insufficient

Conventional digital services frequently rely upon broad privacy policies that authorize collection, product improvement, analytics and sharing with vaguely described partners. Users are expected to agree before they can access the service.

This model is already weak for ordinary personal information. It becomes unacceptable for neural data.

Consent must identify the category of neural information, the purpose of processing, the intended recipients, the duration of retention and whether the information will be used to train artificial intelligence. Permission for medical treatment should remain separate from permission for marketing, commercial research or unrelated product development.

Users should also be told which outputs are generated from their signals. A company that claims to collect only “performance data” should disclose whether that category includes predictions about attention, emotion or cognitive capacity.

Most importantly, consent must be revocable. A person should not surrender permanent rights over future interpretations of his nervous system by accepting a product once.

The Emerging Legal Recognition of Neural Data

Law is beginning to recognize that neural information requires specific protection.

Colorado enacted legislation in 2024 expanding the Colorado Privacy Act’s definition of sensitive data to include biological data and neural data. The legislation defines neural data as information generated through measuring activity in the central or peripheral nervous system that can be processed by or with the assistance of a device.

California subsequently amended its consumer privacy framework to include neural data within sensitive personal information. These developments are significant because consumer neurotechnology may fall outside protections traditionally associated with medical records.

At the international level, the OECD’s Recommendation on Responsible Innovation in Neurotechnology calls for the protection of personal brain data, informed consent, access and deletion options, anti-discrimination safeguards and anticipation of misuse. Its July 2025 Neurotechnology Toolkit supports implementation of those principles.

UNESCO’s April 2025 draft Recommendation on the Ethics of Neurotechnology also addresses mental privacy, cognitive liberty, neural data, informed consent and the right to refuse or withdraw from neurotechnology. As of this article’s publication, it remains a proposed international instrument rather than an adopted global standard.

These initiatives establish an important direction, but legal classification alone cannot produce cognitive sovereignty. Rights must be implemented through technological architecture.

A Neuro-Cypherpunkist Framework for Neural Data

The Neuro-Cypherpunkist Manifesto declares that neural information belongs fundamentally to the individual from whom it originates. Translating that declaration into practice requires the following safeguards.

Sensitive by Default

Identifiable neural data should be classified as sensitive by default. Lower-risk treatment may be justified for narrowly processed, anonymous or transient outputs, but institutions should bear the burden of demonstrating why weaker protection is appropriate.

Local Processing

Raw neural signals should be processed locally whenever technically practical. External services should receive only the minimum command or result required to perform the function selected by the user.

Purpose Limitation

Information collected for treatment, rehabilitation, research, entertainment or device control should not be repurposed without a separate and legitimate basis. A therapeutic recording must not quietly become advertising intelligence.

Inference Protection

Protections must apply to cognitive and behavioural inferences as well as raw recordings. Institutions should disclose consequential inferences and allow individuals to correct or contest them.

Encryption and Authentication

Neural information must be protected during collection, storage and transmission. Connected devices should authenticate commands and updates, with particular safeguards around systems capable of stimulation.

Separation of Read and Write Authority

Access to neural recordings must never automatically grant the ability to transmit stimulation or alter device behaviour. Read and write privileges should be technically separated and independently authorized.

Individual Access and Portability

Users should be able to access their neural information, understand how it has been processed and obtain usable copies where appropriate. Learned profiles required for essential functions should not become instruments of provider lock-in.

Restricted Commercial Exploitation

Neural data should not be sold, used for targeted psychological advertising or supplied to data brokers without specific and informed authorization. Consent to use a device is not consent to enter a market for mental information.

Protection from Discrimination

Employers, insurers, schools and governments should not make consequential judgments from neural information without a legitimate purpose, demonstrated reliability and effective safeguards against discrimination.

The Right to Delete—and the Duty to Preserve When Necessary

Individuals should generally be able to delete information no longer required for its authorized purpose. Medical safety, legal accountability and valid scientific research may justify limited retention, but these exceptions must be clearly defined and securely governed.

The Right to Refuse

No person should be compelled to surrender neural information merely to participate in ordinary employment, education, commerce or public life. Refusal must remain practical rather than nominal.

My illustration “Memories Under Glass” work-in-progress. The art represents brain data exposing a person’s inner life, not merely their online activity.

Neural Data and Artificial Intelligence Training

Neural information may become especially valuable for training artificial intelligence. Large datasets can help models recognize intended movement, attempted speech and patterns relevant to neurological conditions.

This research may create enormous public benefit. It could improve assistive devices, accelerate medical discovery and make interfaces more accurate for a wider range of users.

But beneficial purpose does not eliminate the need for consent and governance. Neural information collected to operate one device should not automatically become training material for unrelated commercial models. De-identification, security, scientific oversight and limits on secondary use remain necessary.

Individuals contributing neural information to research should receive an understandable explanation of whether models, features or commercial products may be developed from it. Where withdrawal of data from a trained model is technically impossible, that limitation should be disclosed before consent is obtained.

Open science and cognitive sovereignty can coexist when research is designed around legitimate permission, minimization, transparency and protection against re-identification.

The Danger of a Neural Data Economy

The digital economy established a model in which services are exchanged for extensive behavioural observation. Platforms collect information, predict preferences and sell access to attention. Users receive convenience while institutions accumulate increasingly detailed profiles.

That model must not be carried unquestioned into neurotechnology.

Neural intimacy platforms could transform cognitive and emotional information into commercial infrastructure. A neural advertising economy might seek to measure not merely what people click, but how their nervous systems respond. Data brokers could attempt to combine neural profiles with health, financial and behavioural records.

As examined in The Brain Brokers, sufficiently valuable neural information could also attract unauthorized markets, theft and coercive acquisition.

The history of the internet shows that commercial practices established during an industry’s early development can become difficult to reverse. Neural privacy must therefore be embedded before surveillance becomes the assumed price of participation.

Neural Data as an Extension of the Person

The strongest protection for neural data does not arise merely because it is medically sensitive, commercially valuable or difficult to replace. It arises because neural information occupies an unusual position between data and personhood.

A neural recording is not the person. An algorithmic model is not the mind. A prediction is not an intention, and neurological activity should never be confused with moral guilt or completed action.

Yet these records originate from and describe processes intimately connected with how the person perceives, communicates and acts. Their collection can affect mental privacy. Their interpretation can affect identity. Their misuse can affect freedom, while their manipulation may affect bodily or cognitive integrity.

Neural data must therefore remain subject to the continuing authority of the individual. Sharing it for a defined purpose should not extinguish that authority. Providing a device should not transfer ownership of cognition to the manufacturer. Agreeing to analysis should not create unlimited permission for future interpretation.

This is the principle of neural self-custody: the individual must retain meaningful authority over the information and technological permissions governing the connection to his nervous system.

The Neuro-Cypherpunkist Position

Neuro-Cypherpunkism does not demand that all neural information remain inaccessible. Medicine requires information. Scientific research requires data. Assistive technologies require calibration, and some advanced systems may require remote computation.

The philosophy demands that access remain justified, proportionate, secure and subordinate to the person.

Neural data is not ordinary data because it may be produced without deliberate expression.

It is not ordinary data because it can be used to infer states that remain internal.

It is not ordinary data because it may precede observable action.

It is not ordinary data because its meaning changes as analytical technology improves.

It is not ordinary data because errors can alter how institutions judge the person.

It is not ordinary data because the same technological system may both observe and influence the nervous system.

It is not ordinary data because some users may depend upon the collecting device for communication, movement or independence.

And it is not ordinary data because it originates from the biological processes underlying human agency itself.

Technology may measure the signal.

Artificial intelligence may interpret the signal.

A device may translate the signal into action.

But none of these capabilities transfers sovereignty over the person who produced it.

Thought is not telemetry.

Cognition is not a commodity.

Neural data is not ordinary data.

The mind is the final private key.