From an intuition to a rigorous frame
In Volume I, neuroethics arrived as a closing caution: brains are special, be careful. That is true but not yet useful. At the graduate level the task is to turn the caution into a specification — a set of properties a BCI must have or preserve, stated precisely enough that an engineer, a clinician and a regulator can argue about whether a given system meets them.
The central claim of this track is that ethics for neurotechnology is not a soft appendix bolted onto the science. It is a design constraint with the same status as latency, power budget, or biocompatibility. A decoder that leaks inner state, a controller that blurs who acted, or a device abandoned after a trial are not merely regrettable — they are engineering failures against a requirement, and the rest of this track gives you the requirements.
What makes neurotechnology ethically distinctive
Not everything about BCI ethics is new; medical devices, wearables and machine learning all raise privacy and safety questions. Four features, taken together, make neurotechnology genuinely distinctive rather than just another data source.
- Directness — the signal is drawn from the organ of thought itself, upstream of the behaviour a person chooses to show. There is less room to strategically self-present.
- Reading and writing — unlike a fitness tracker, an implant can both infer state and stimulate the brain, so the device can in principle change the very mind it is measuring.
- Continuity with the self — the brain is the substrate of memory, personality and agency, so failures here threaten identity, not just data or a limb.
- Vulnerable populations first — the people who benefit most (paralysis, ALS, locked-in) often have the least ability to consent, refuse, or walk away, which sharpens every other concern.
The moral vocabulary: rights, principles, capabilities
Three frameworks recur, and you should be able to switch between them. Principlism (autonomy, beneficence, non-maleficence, justice) is the working language of clinical ethics boards and is excellent for individual cases. A rights framing — the neurorights proposal — asks whether brain data and brain integrity deserve protections strong enough to name explicitly in law. A capabilities framing asks not 'was a rule broken?' but 'what is this person now able to do and to be?', which is often the right lens for access and equity.
These are lenses, not rival teams. A mature analysis of, say, a consumer EEG product that sells aggregate 'attention' data will invoke autonomy (did the user meaningfully agree?), a right to mental privacy, and a capabilities question (does cheap neurotech widen or narrow real opportunity?). Fluency means reaching for whichever lens exposes the problem most clearly, not defending one school.
Reading versus writing: a taxonomy of interventions
The single most clarifying axis is direction of information flow. Reading systems infer state (a P300 speller, a motor decoder, a passive BCI monitoring workload). Writing systems inject state (deep brain stimulation, sensory microstimulation, closed-loop neuromodulation). Bidirectional systems do both in a loop, and it is the loop that generates the deepest problems in Guide 3.
The ethical weight is not symmetric. Reading primarily threatens privacy and autonomy of information; the harm is disclosure or inference. Writing threatens integrity and authenticity of the person; the harm is altered mood, personality, or will. Regulators feel this asymmetry too: a stimulation device that can change affect is scrutinised differently from a passive recorder — though, as Guide 5 shows, consumer 'reading only' products slip through exactly because they look harmless.
The road map: four hard questions and a governance layer
The rest of the track is organised around four questions and the machinery each demands. Privacy (Guide 2): what can actually be read, and what is the formal toolkit that limits leakage? Agency (Guide 3): when the decoder acts with you, who is the author, and who is responsible? Identity and consent (Guide 4): can an intervention change who you are, and what does consent mean for people who cannot easily give it? Governance (Guide 5): given all of the above, how are rights, laws and regulators actually responding?