health · digital · family: the rulebook was written for a world that no longer exists
the implant outlives the company
Brain-computer interface patients face abandonment when manufacturers fail — and no regulatory framework prevents it
Problem statement
Implantable brain-computer interfaces (iBCIs) are emerging medical devices that create direct communication pathways between the brain and external technology. Unlike traditional medical implants, iBCIs often rely on continuous software updates, cloud-based systems, and technical maintenance to function. If the manufacturer goes bankrupt, pivots its business, or simply stops supporting the product, the patient may be left with a non-functional or degrading implant in their brain with no pathway for continued care. The FDA's current regulatory framework does not require manufacturers to establish contingency plans for device support, data handover, or explantation funding in the event of company failure. As companies like Neuralink advance iBCI technology, this gap will affect an increasing number of patients.
Why this matters
Current patient numbers are small (roughly 60 people had received BCI implants as of October 2024), but BCIs are moving from research toward clinical and commercial use. Each affected patient faces the risk of carrying a non-functional implant whose surgical removal carries risks similar to implantation, including potential damage to brain tissue, or whose components can break down within the brain, causing inflammation or other adverse reactions. BCI developers include start-ups and privately backed companies like Neuralink, and start-up sponsors have failed before, yet patients who receive implants during clinical trials or early commercialization are vulnerable because the device depends on manufacturer-provided software and support. The safeguards set now, before adoption scales further, will shape the protections for the next generation of BCI users.
What’s been tried and why it hasn’t worked
The FDA's 2021 BCI guidance provided recommendations for nonclinical testing and IDE study design but did not address long-term manufacturer support obligations, software dependency management, or patient abandonment scenarios. Cases have already emerged of patients with abandoned neurotechnology implants who lost device functionality when a company failed (the Argus II retinal prosthesis, after Second Sight collapsed in 2020) or lost the device itself when a study ended and funding ran out; the risks of abandonment the Knowing Neurons paper names include medical complications and data security vulnerabilities. The FDA oversees pre-market approval and limited post-market monitoring but does not require companies to provide ongoing software support, device maintenance, or removal options once a product is discontinued. No system currently exists to ensure patient protection once industry or research support ends. Surgical explantation carries its own risks, similar to implantation, while a deactivated device left in place may have long-term health implications, such as an inability to safely get an MRI or complications of device migration.
What would unlock progress
A regulatory framework requiring BCI manufacturers to establish contingency plans before market authorization — including source code escrow, encryption key handover procedures, minimum support duration commitments, and financial provisions for explantation. An industry consortium model (similar to nuclear decommissioning funds) where BCI manufacturers pool resources to guarantee patient support. Open-architecture BCI designs with standardized interfaces that could be maintained by third parties if the original manufacturer fails. The precedents the Knowing Neurons paper describes — the Argus II retinal prosthesis, left unsupported when its maker collapsed, and BrainGate, whose research was continued by academic institutions after its start-up sponsor ceased operations — provide both cautionary examples and potential models for continuity frameworks.
Entry points for student teams
A student team could draft a model regulatory framework for BCI manufacturer continuity obligations, drawing on precedents from nuclear decommissioning funds, pharmaceutical orphan drug requirements, and software escrow agreements. Another entry point would be designing an open-architecture BCI firmware platform with standardized APIs that would allow third-party maintenance if the original manufacturer ceases operations. A policy/engineering hybrid team could analyze current informed consent documents from BCI clinical trials and identify gaps in communicating long-term software dependency risks. Teams with backgrounds in regulatory science, bioethics, embedded systems, or health policy would be well-suited.
Genes
Structural cousins: stuck the same way, in other fields
Sources
FDA Guidance, "Implanted Brain-Computer Interface (BCI) Devices for Patients with Paralysis or Amputation — Non-Clinical Testing and Clinical Considerations," 2021, supplemented by Knowing Neurons, "The Aftermath of Abandoned Neurotech" (2025); Frontiers in Human Neuroscience, "Mind the gap: bridging ethical considerations and regulatory oversight in implantable BCI human subjects research" (2025). Accessed 2026-02-19. go to source ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
- FDA Guidance on Implanted BCI Devices (2021) focused on biocompatibility, electrical safety, sterility, and clinical study design — none of the identified gaps regarding long-term software support or manufacturer failure contingencies were addressed.
- Knowing Neurons (2025) documented real cases of patients with abandoned neurotechnology implants. Frontiers in Human Neuroscience (2025) identified regulatory gaps in IRB and FDA oversight of implanted BCI human subjects research.
- Neuralink is advancing its BCI in people with paralysis under FDA-authorized clinical trials, without full market approval; its devices are implanted under investigational exemptions. According to the Knowing Neurons paper, private neurotechnology firms like Neuralink operate without obligations for long-term patient care.
- The "window" temporal tag reflects that most iBCIs are still investigational — establishing continuity frameworks now (before adoption scales further) is far more tractable than retrofitting them later.
- Related to broader patterns of software-dependent medical devices where the device lifecycle exceeds the manufacturer's software support cycle.
Reconciliation 2026-09-30: The whole write-up was read against the three sources it names, all fetched on 2026-09-30 and read in full. The FDA guidance was read from the FDA page on the Source line and from its PDF (https://www.fda.gov/media/120362/download, issued May 20, 2021). The Knowing Neurons paper was read at https://knowingneurons.com/abandoned_neurotech/ (by Elizabeth Bottorff, dated May 8, 2025); its editor's note says it was corrected in April 2026, after this write-up was made, so an Internet Archive copy from September 2025 was compared with it. The Frontiers in Human Neuroscience paper (Wilkins, Coffin, Pham, Klein and Marathe, volume 19, article 1633627, 2025, https://doi.org/10.3389/fnhum.2025.1633627) was read from the open copy at Europe PMC (PMC12325254). Changes: (1) The Source line and Source Notes dated the Knowing Neurons paper 2024; the page is dated May 8, 2025, and both now say 2025. The Frontiers title on the Source line was cut short; it now ends "in implantable BCI human subjects research," as in the record. (2) The text said that "unlike traditional implants such as pacemakers," iBCIs depend on "continuous software support, cloud connectivity, firmware updates, and algorithmic calibration." The Knowing Neurons paper says that, unlike traditional medical implants, neurotechnologies often rely on continuous software updates, cloud-based systems, and technical maintenance. It does not mention pacemakers, firmware or calibration, and it gives a heart implant (St. Jude, 2017) as an example of lapsed support. The sentence now follows the paper. (3) "The patient is left with" a non-functional implant is now "may be left with," which is the strength both papers give it. (4) Synchron, Paradromics and Blackrock Neurotech are not named in any of the three sources; only Neuralink is. The three names were taken out. (5) "Dozens in clinical trials" is now the paper's figure: roughly 60 people had received BCI implants as of October 2024. "The market is projected to grow rapidly as commercial authorizations are obtained" is not in the sources; the sentence now says BCIs are moving from research toward clinical and commercial use. (6) The text said an implant "cannot be safely explanted without risk of brain damage" or "degrades in ways that affect neural function." The Frontiers paper says surgical explantation carries risks similar to implantation, that brain surgery can damage brain tissue, and that implant components can break down within the brain, causing inflammation or other adverse reactions. The sentence now says that, keeping the paper's hedges (potential damage; components can break down). (7) "The BCI industry follows the venture-capital startup model where company failure rates are high" is not in the sources. The Knowing Neurons paper says Neuralink is largely backed by private capital and describes start-up sponsors that failed. The sentence now says that, naming Neuralink as the paper's example of a privately backed company; "uniquely vulnerable" is now "vulnerable," and "cannot function without" is now "depends on." (8) "Thousands of future patients" and "before thousands of patients are implanted" are not in the sources; both now follow the paper's wording about acting before adoption scales further. (9) The text said patients "lost device functionality when research teams dissolved or companies failed, facing medical complications and data security vulnerabilities." The paper's cases are the Argus II retinal prosthesis (users left with non-functional implants when Second Sight collapsed in 2020) and a BCI research participant whose working implant was removed in 2021 when his study ended and funding ran out. It names medical complications and data security vulnerabilities as risks of abandonment, not as things that happened in those cases. The sentence now says that, and names the Knowing Neurons paper as the source of the list of risks. The April 2026 correction to the paper changed its account of the BrainGate trial: participants faced the risk of unusable implants when Cyberkinetics ceased operations, and academic institutions continued the research. (10) "The FDA lacks explicit statutory authority to mandate post-approval manufacturer support obligations" is not in the sources. The paper says the FDA oversees pre-market approval and limited post-market monitoring but does not require companies to provide ongoing software support, device maintenance, or removal options once a product is discontinued. The sentence now says that. (11) Two sentences were removed because no source makes either claim: that requiring financial escrow or source code deposit "would be unprecedented in device regulation," and that "the startup ecosystem resists binding long-term obligations that increase burn rate and reduce investment attractiveness." (12) "No insurance or government program exists to fund explantation or continued device support" is now the paper's statement that no system currently exists to ensure patient protection once industry or research support ends. (13) "The surgical risks of explanting a chronically implanted brain device may exceed the risks of leaving it in place" is not in the sources. The Frontiers paper says explantation carries risks similar to implantation and that a deactivated device left in the body may bring long-term problems, such as an inability to safely get an MRI or complications of device migration. The sentence now says that. (14) The cochlear implant precedent ("where market consolidation and product discontinuation are common") is not in any source. It was replaced with the two precedents the Knowing Neurons paper does describe, Argus II and BrainGate (Second Sight ceased support in 2020; academic institutions continued the BrainGate research after Cyberkinetics ceased operations). (15) In Source Notes, "Neuralink received FDA Breakthrough Device Designation and initiated human trials in 2024," "Synchron has enrolled patients in a pivotal trial for its endovascular BCI," and "No manufacturer has publicly committed to binding long-term support obligations" are not in the sources. The bullet now gives what the Knowing Neurons paper says about Neuralink: FDA-authorized clinical trials, no full market approval, devices implanted under investigational exemptions, and no obligations for long-term patient care. (16) In Source Notes, "the BCI industry is still in early commercialization" is now "most iBCIs are still investigational," which is the Frontiers paper's statement. Confirmed in the sources and left as written: the long title and the display title; the FDA guidance title and year; that the guidance gives recommendations for non-clinical testing and for the design of feasibility and pivotal studies, covers biocompatibility, sterility and electrical safety, and says nothing about manufacturer failure, continued support or abandonment; that the FDA does not require contingency plans; the proposal that manufacturers set contingency plans before approval, with minimum support durations, data handover procedures and financial provisions for explantation; and that the Frontiers paper describes gaps in IRB and FDA oversight. Not checked against a source, because they are the write-up's own proposals and not claims the sources make: source code escrow, encryption key handover, the industry consortium modelled on nuclear decommissioning funds, open-architecture designs, the student entry points, and the last bullet above on software-dependent devices in general. A second reader went over every repair on the same day (2026-09-30) against the same three sources, fetched independently (the live Knowing Neurons page and its September 2025 Internet Archive copy, the Europe PMC full text of the Frontiers paper with the Crossref record for its title, and the FDA landing page and guidance PDF), and confirmed each repair in the source passage it rests on; no repair was undone or reworded. The second reader also checked that none of the four company names other than Neuralink, the cochlear implant precedent, the venture-capital and escrow claims, or the figure of thousands of patients appears in any of the three sources (the FDA guidance's two hits for "Synchron" are the word "synchronous" in its output-channel table, and its one "pacemaker" is a study exclusion criterion, not a contrast with BCIs). Genome tags untouched.