We expect China to lead everyday use of brain implants through 2030, because it can pay for them. US implants will likely win approval but stall at the payment stage.
Key points
- By the end of 2030, we think China will have more people using a commercially implanted BCI outside a trial than the US, and that no US implant will have a Medicare national coverage determination (about 65%).
- The deciding factor is a buyer, not engineering: China's state insurer coded its first implant within days, and Medicare has no coverage framework.
- US speech neuroprostheses have the strongest evidence of real-world value, but that evidence rests on too few people to satisfy payers soon.
- Consumer earbuds will read brain signals at far larger scale than implants, under consumer rules rather than health privacy law.
- Anyone working in this field should plan for the payment decision and judge progress by coverage, not by the number of approvals.
Our forecast
By 31 December 2030, China will have more patients living with a commercially implanted BCI, used outside a clinical trial, than the United States, and no implanted BCI will hold a Medicare national coverage determination. We put this at about 65%. The two halves support each other. If there is no US coverage path, US commercial use stays limited to self-paying patients and trial extensions.
Why we think so
Every earlier part of this series came back to the same gap: the devices are arriving before the systems that pay for them and govern them. In our signal scan, the hardware reached the market first. Both countries now have a path to approval. Only China has a path to payment. NEO received an insurance code within days of approval, the same day BCI became a Five-Year Plan priority1. Behind it are 134 registered trials and five approved products2. In the US, Synchron has only just funded the pivotal trial that would support the first premarket approval filing3. Even after approval, a former CMS chief medical officer warns that trials built around cursor control may not meet Medicare's health-outcome test4.
Brain Implants Can Win Approval Faster Than They Can Prove Value explains why US approval still will not bring broad use. The best proof of value is one man with ALS who has used a speech implant at home for 3,800 hours and kept a full-time job5. Paradromics' results come from company releases, and its follow-up runs six years6. Neuralink's production targets come from social media posts, not peer review7. Payers will want more than that.
Brain Interfaces Meet Voice Rights, Chinese Robots and AI Labs showed that China is applying a method it has already used. For humanoid robots it named a priority, wrote national standards, then scaled to about 97% of global shipments89. In What Changes When Brain Implants Become Billable Products, we argued that once the buyer exists, volume follows, and so do the pooled neural datasets that make the next decoder cheaper to calibrate10.
The consumer side moves faster. The same company sells an FDA-cleared in-ear EEG and an uncleared consumer earbud11. Short consumer recordings can identify the wearer12, and Vermont removed consent and penalties from its law before passing it13. Implants will remain rare. Brain sensors in earbuds will not.
How this call fails
Suppose it is 2030 and we were wrong. The most likely cause is that CMS acted early. Synchron files with function-based endpoints, the FDA and CMS align their reviews as Fleisher urged4, and a coverage decision follows approval within about a year. US volume then rises quickly, especially if Neuralink's automated surgery works at scale7.
The second path to failure runs through China. NEO's modest functional gains may fail to persuade hospitals, or a durability problem may surface. Longevity is the barrier China's own landscape study names first21. In that case uptake stays in the hundreds, and the US overtakes China through self-pay patients alone.
Who loses
- Disabled US Medicare beneficiaries, the core eligible group, who face self-pay prices for devices their counterparts in China can get reimbursed4.
- US implant makers whose valuations assume approval means revenue, when payment arrives years later3.
- Earbud users, children among them, whose neural data falls under FTC rules rather than HIPAA11.
First moves
- Device makers: add communication and employment outcomes to pivotal endpoints now. Harrell's job is the benchmark54.
- Payers and CMS: publish a draft coverage framework before the first PMA filing, not after approval.
- State legislators: restore consent requirements, private rights of action and penalties before Vermont's 2028 consent rule takes effect13.
- Employers and insurers: treat neural data as a biometric identifier, not anonymous telemetry1214.
- Investors: weigh coverage decisions and multi-participant peer-reviewed results above approval counts.
What would change our mind
- CMS proposes a coverage pathway or national coverage determination for implanted BCIs before Synchron's pivotal readout3.
- Chinese commercial implant numbers stay below a few hundred by 2028, or there is a NEO safety recall1.
- Peer-reviewed speech results from several participants across at least two US device makers by 202767.
- A federal neural-data bill with consent and penalties advances out of committee13.
Sources
- China's NMPA grants world's first commercial approval for an invasive BCI and assigns it a health-insurance code within days · journalism ↩
- Landscape study counts 134 registered Chinese BCI trials and five NMPA-approved products, 79% of them non-invasive · peer-reviewed ↩
- Synchron raises $200M to run the first pivotal trial aimed at an FDA premarket approval for a permanently implanted BCI · journalism ↩
- Former CMS chief medical officer warns no Medicare coverage framework exists for implanted BCIs nearing approval · expert ↩
- Nature Medicine reports ~2 years and 3,800 hours of independent at-home use of an AI speech neuroprosthesis at 99% word accuracy · peer-reviewed ↩
- Paradromics completes first permanent Connexus implant and reports open-ended real-time speech within three months · journalism ↩
- Neuralink targets 'high-volume' implant production and near-automated surgery in 2026 as enrollment passes 20 across four countries · journalism ↩
- China reaches volume production and publishes a national standard system for humanoids · institutional ↩
- China's MIIT projects >100,000 humanoids produced in 2026; two firms take ~80% share · institutional · also trendforce.com ↩
- PNPL 2026 benchmark pushes non-invasive speech decoding toward word classification with 10 minutes of per-user data · peer-reviewed ↩
- FDA clears first in-ear EEG device while the same company ships a non-medical consumer version by year-end · journalism ↩
- Consumer EEG headset can identify individuals from short recordings regardless of emotional state, study finds · peer-reviewed ↩
- Vermont becomes fifth US state with a neural-data statute but strips consent requirement and private right of action before passage · journalism ↩
- UNESCO adopts first global neurotechnology ethics standard; legal scholars argue consent and property frames cannot settle who owns neural data · institutional ↩