Brain recordings can support care and research. Patients should also receive clear explanations of what is collected, who can use it and how decisions about future use will be made.
Consider a patient referred for an electroencephalogram, or EEG. Sensors on the scalp record electrical activity in the brain to help investigate a clinical problem, such as suspected epilepsy. [1] The patient may understand why the test is needed without knowing much about the recording’s subsequent use.
If a copy is later proposed for research or for developing an AI system, a new set of questions arises. Who will receive it? Will it remain linked to the patient? What will the new analysis attempt to infer? Trust depends on making those questions visible while there is still an opportunity to discuss them.
Be precise about what the technology reveals
Neural data includes recordings of nervous-system activity. The raw signal, a clinician’s interpretation and a prediction generated by an algorithm are distinct kinds of information. A prediction should be identified as a prediction, with its uncertainty and intended use made clear.
An ordinary EEG is not a transcript of someone’s private thoughts. Research into decoding more complex mental content operates under particular conditions. In a 2023 study, Jerry Tang and colleagues reconstructed aspects of language meaning from functional MRI recordings. The researchers found that cooperation was needed both to train and to apply their decoder. [2]
That result is scientifically significant, but it does not demonstrate unrestricted mind reading from a routine hospital test or a consumer headset. Public discussion needs to distinguish the recording method, the task and the conditions under which a system was evaluated.
Privacy concerns nevertheless deserve attention without waiting for more powerful decoding. A health-related inference may affect someone if another person treats it as reliable. An inaccurate label can also cause harm. The quality of an inference and the permission to use it are separate questions.
Turn rights into understandable choices
In an influential 2017 paper, Marcello Ienca and Roberto Andorno proposed protections centred on cognitive liberty, mental privacy, mental integrity and psychological continuity. Their argument concerned people’s freedom to make decisions about interventions affecting the mind and protection against intrusive or harmful uses of neurotechnology. These were proposals for developing the rights framework, not a description of uniformly enforceable rights everywhere. [3]
UNESCO’s Recommendation on the Ethics of Neurotechnology, adopted in November 2025, emphasises consent, privacy, limits on data use, non-discrimination and accountability. It calls for implementation within national legal systems; it does not establish identical domestic remedies everywhere. [4]
For patients, these principles become useful when explanations address the actual service. The clinical purpose should be clear. Any proposed research, product development or commercial use should be described distinctly, with the relevant consent or other lawful basis established. The explanation should identify who can access the information, how long it is retained and how questions or complaints can be raised. [4]
Choices about withdrawal also need an honest explanation. Before asking for agreement, a service should explain what can be stopped or removed and any limits that apply to information already used. Consent materials should be understandable in the patient’s language and accessible to people with different communication needs. [4]
Follow the recording through the service
A practical proposal for hospitals is to map the journey of a recording from collection to storage, reporting and any onward transfer. At each stage, identify the responsible organisation and the purpose of access. This makes it easier to discover a gap between what a consent form says and what the service actually does.
For example, a hospital considering an external analysis platform could ask the supplier to demonstrate how access is controlled and how its own staff’s access is recorded. It could also ask for a patient-facing explanation of whether recordings contribute to further model development. These questions should be answered before routine use begins.
Patients and caregivers should help test those explanations. Asking them to describe, in their own words, what happens to the recording may reveal ambiguity that a technical review misses.
Responsible use of neural data should support useful care and research while preserving people’s ability to understand and question what is being done. That responsibility continues after the electrodes are removed.
References
- NHS. EEG electroencephalogram. Patient information. Accessed 1 October 2026. Read source
- Tang J, LeBel A, Jain S, Huth AG. Semantic reconstruction of continuous language from non-invasive brain recordings. Nature Neuroscience. 2023;26:858–866. Read source
- Ienca M, Andorno R. Towards new human rights in the age of neuroscience and neurotechnology. Life Sciences, Society and Policy. 2017;13:5. Read source
- UNESCO. Recommendation on the Ethics of Neurotechnology. Adopted 11 November 2025. Read source
