At Science Corporation, we’re working to advance brain-computer interface (BCI) technology and believe the world will soon see major breakthroughs in brain-related medical treatments. Decades of research and engineering are converging into advances previously only imagined.
Our optimism about what we hope to do for millions of people with serious unmet medical needs comes with a deep responsibility to engage with the broader implications of our work – and our duty to patients and the world.
As we launch new products aimed at making neural engineering technology readily available to researchers, we’ve been studying the trade-offs that get made whenever a neural interface system is designed, especially the tension between privacy and security on the one hand, and usability and simplicity on the other. The technologies we build must embody our philosophy and make these tradeoffs, when necessary, in a clear, principled, and explainable way.
As Chief Legal Officer at Science, I see firsthand how making these product and engineering decisions can raise complex ethical questions. And so it seems important to outline some of our commitments to the people we ultimately work for: the patients whose quality of life we seek to improve.
Fundamental agency and privacy
We believe that all patients using our devices should have agency over their lives. To this end, the devices we build must in their basic structure prioritize patient rights, such as privacy and ownership, and in doing so respect each patient’s desires for their treatment and broader life goals.
More concretely, we believe this commitment means that people must have maximum insight into and control over the inputs, processes, and outputs of the devices so directly connected to the functioning of their mind.1
For example, patients’ privacy rights in their neural data must assure that data transfers are free from unwanted or unknown surveillance.2 By extension, this means that a BCI must function free of undesired or unexplained monetization models (“to continue using your neural implant, please enjoy this 30-second ad for breakfast cereal”), or obscured security tradeoffs (“we’re excited to offer a free trial facial recognition upgrade from our trusted AI partner”).
We recognize that data flows and security processes are often complex and must be carefully and thoroughly articulated and explained in order to be well understood by anyone considering or using a BCI.3
Full device functionality offline
We believe that ensuring consistent patient experience, usability, and quality of life means that neural interface devices’ core features must work even without a network connection.4
This commitment represents a significant design challenge in a world where devices are increasingly always online. But we believe offline functionality is non-negotiable because it ensures that any technology we build rests on simple foundations that will work even when other, related systems fail (as they inevitably will), and that responsibility for the device’s core functionality rests with us rather than being widely dispersed among others.
In line with graceful degradation and progressive enhancement lessons learned from two decades of web development, network access should enhance functionality but not be required for it.5 Ideally, a device should work to a level of minimal functionality anywhere and in any condition a patient might find themself, such as extreme or unusual environments, and in disaster situations with unexpected constraints on normal resources, such as internet or power. Or, barring that, the device’s loss of functionality under unusual conditions should leave the patient no worse off for having it.
Offline functionality enhances user control of certain data inputs or outputs, such as geolocation data, and will allow device operation in constrained or specialized environments, such as around other medical equipment affected by WIFI signals, or in highly secure environments.
Patient ownership of their devices
Over the last few decades, we’ve increasingly seen longstanding norms around bedrock property rights whittled away and replaced by limited licenses, recurring subscriptions, and other kinds of relatively weak and ephemeral rights in what we buy.6 With this trend, crucial rights related to ownership – such as the right to repair or modify what you own – have correspondingly deteriorated.7
We believe that BCIs must resist and stand apart from this broader commercial trend. A neural device purchased must fully belong to its owner.
This means that the device’s full functionality must never be gated by a subscription, manufacturer’s conditions (such as allowing advertising or surveillance), proprietary add-ons or other forms of partial or contingent access or control. Continuing core functionality cannot depend on continued payment, and available features should not be withheld or weakened to extract greater payment from patients as a business model.
And while modern software development is dynamic and relies on constant improvement and updating, we believe that ownership of neural devices must include the right to control how and by whom they are repaired or modified. Nothing should stand in the way of your neural interface being yours and serving your needs.
When tradeoffs between functionality and privacy/security inevitably arise, our aim should be that whenever and to the greatest extent possible, the patient has control over such tradeoffs in line with their preferences.
More concretely, this means that a neural device user should never be presented with false, forced, or misleading choices that benefit any other party, including the manufacturer, a commercial third party, or the government. In practice, data that is necessary for device functionality must be distinctly presented to the user and not forcibly or misleadingly bundled with data transfers or other actions that are merely helpful or serve unarticulated mixed purposes.
Additionally, in the future, a full-featured and integrated BCI may raise some ambiguity about where the patient ends and the device begins.8 This in turn raises philosophical questions, but in the context of ownership and surveillance, one in particular comes to mind – a person’s BCI and neural data should not be used against them, in court, investigations, or other situations where putative insight into a person’s cognition, or other data, might be used against them.9 The current political moment shows parallel efforts to roll back ownership rights in the goods we purchase and rights of control over our bodies. BCI developments underscore the importance of these rights and are a potential bulwark against such trends.
Going forward
We expect that as our work evolves, so will these commitments. We believe that neurotechnology is on the cusp of important breakthroughs that will significantly improve and transform human life. We are determined to help lead that effort, and to ensure patients have the best and most complete access possible to innovative technologies that address their condition. As we learn more as a company and industry, we are committed to challenging our existing assumptions and developing new insights about patients’ rights and our responsibilities.
The commitments outlined above are meant as a starting point. We welcome comments, ideas, and contributions from others in our field, and from stakeholder groups, especially patients, regulators, and ethicists.
- See, generally, Nita A. Farhany, The Battle for Your Brain: Defending the Right to Think Freely in the Age of Neurotechnology (St. Martin’s Press 2023).
- Jwa AS, Poldrack RA. Addressing privacy risk in neuroscience data: from data protection to harm prevention. J Law Biosci. 2022 Sep 4;9(2).
- F. Brocal, Brain-computer interfaces in safety and security fields: Risks and applications, Safety Science 160 (2023).
- For a background discussion on the challenges of implantation, see The Royal Society, iHuman: Blurring Lines Between Mind and Machine, available at https://royalsociety.org/ihuman-perspective
- https://www.w3.org/wiki/Graceful_degradation_versus_progressive_enhancement
- A representative article on how digital goods affect the experience of ownership is, Watkins, R. D., Denegri-Knott, J., & Molesworth, M. (2015). The relationship between ownership and possession: observations from the context of digital virtual goods. Journal of Marketing Management, 32(1–2), 44–70. And for background on the underlying property concepts, I recommend Stephen Munzer, A Theory of Property, Cambridge UP (1990).
- On the right to repair, see Aaron Perzanowski, The Right to Repair: Reclaiming the Things We Own (Cambridge UP, 2002).
- Trimper JB, Wolpe PR, Rommelfanger KS. When “I” becomes “We”: ethical implications of emerging brain-to-brain interfacing technologies. Front Neuroeng. 2014 Feb 12;7:4. doi: 10.3389/fneng.2014.00004.
- For discussion of the applications and temptations of technology-enabled “mindreading” see Henry T. Greely, Neuroscience, Mindreading, and the Courts: The Example of Pain, 18 J. Health Care L. & Pol’y 171 (2015). Available at: http://digitalcommons.law.umaryland.edu/jhclp/vol18/iss2/2
