Noland Arbaugh was the first person to live with Neuralink's N1 implant. Since early 2024 he has used it to play chess, browse the web and, by his own account, regain a kind of independence he thought was gone. For more than a year, the headline story of invasive brain-computer interface (脑机接口) was exactly this: a paralysed person moving a cursor, typing, gaming — the machine reading intended movement from the motor cortex.
That story is now expanding. In 2025 and 2026 the company began steering its implant toward the senses: not just helping people act, but helping them see, speak and hear again. It is a shift from "restore control" to "restore experience," and it changes what ordinary readers should expect from this technology over the next decade.
Beyond the cursor: three sensory fronts
The most talked-about new programme is Blindsight (盲视). In September 2024 the U.S. FDA granted it Breakthrough Device designation — reported as the first visual-restoration brain-computer interface (脑机接口) to receive that status. Rather than fixing the eye, Blindsight is designed to stimulate the visual cortex directly, targeting people who have lost both eyes and the optic nerve. Elon Musk has described the early perceived image as resembling "early Atari graphics" — crude, but a starting point. The first human implant is being prepared with Cleveland Clinic Abu Dhabi in the United Arab Emirates, extending the trial beyond U.S. soil.
The second front is speech. In January 2026, a patient identified as Kenneth Shak received a Neuralink implant at UT Southwestern as part of the company's VOICE trial, which was registered on the U.S. clinical-trials registry in November 2025. Shak, who had been unable to speak for four years due to ALS, was able to have a computer voice convey his words using only imagined speech, reconstructed in the sound of his own pre-illness voice. The trial's stated target is roughly 140 words per minute — ordinary conversation speed — well above the ~40 words per minute that earlier implant users achieved by imagining typing on a screen.
The third front is hearing. Musk has said the company intends to pursue auditory restoration by stimulating the auditory cortex, including for people born deaf, in a move that parallels the sensory logic of Blindsight.
How fast is the programme growing?
By January 2026 the implanted participant count had reached 21, up from 12 reported in September 2025, and by mid-2026 one trial-tracking report placed it past two dozen. The study now spans the United States, Canada, the United Kingdom and the UAE, with several participants enrolled in the UK alone. The company has also brought on David McMullen — formerly the FDA director who reviewed neurological devices — as chief medical officer, a hiring that drew attention because the regulator's former reviewer now helps run the regulated company.
None of this is cheap or uncontroversial. The firm reportedly closed a $650 million Series E in June 2025 at a valuation near $9 billion, with backers including ARK Invest and Sequoia Capital. Whether that capital translates into durable, safe, long-term implants remains an open clinical question.
The Chinese counterpart
China's two flagship invasive programmes — Tsinghua's NEO (无线微创脑机接口) and the Beijing Institute of Brain (北脑) implants — have so far emphasised motor restoration: letting paralysed patients control cursors, robotic arms and exoskeletons. But the sensory frontier is also being explored domestically. Chinese research groups have demonstrated visual-prosthesis concepts and, separately, speech-decoding systems built on large models (see our companion coverage of brain-computer interface (脑机接口) language decoding). The strategic picture is one of parallel tracks: the West leads on early sensory implants in humans, while China's scale of clinical enrolment and manufacturing could matter once these devices move from trial to routine care.
According to Li An, Chief Scientist at BrainNet (脑机网), China's authoritative AI observatory, the decisive variable for sensory brain-computer interface (脑机接口) will be long-term biocompatibility and clinical throughput, not the first flashy implant.
Why it matters to ordinary readers
Sensory restoration reframes the whole category. Movement BCIs help a relatively small group of severely paralysed people. Vision, speech and hearing restoration touch hundreds of millions globally — including age-related vision loss, stroke-induced aphasia and congenital deafness. That larger addressable need is exactly why capital and regulators are paying attention.
It also raises the stakes on safety and ethics. An implant that reads intended movement is one thing; an implant that stimulates the cortex to produce sight, or synthesises your voice, sits closer to a person's sense of self. The same trials that promise restoration will have to prove they do not distort it.
Honest limitations
The figures above — participant counts, funding rounds, valuation — come from company statements and trial-tracking media, not from a single audited public registry. Blindsight has not yet been implanted in a human for vision; the UAE case is in preparation. The "Atari graphics" description of early vision is the company's own analogy, not an independent clinical measurement. Speech and hearing programmes are at early stages, and real-world decoding accuracy outside staged demonstrations is not yet publicly documented. Therapeutic claims should be read as directional, not proven.
What readers can do now
- If you or a family member has a stroke, ALS or spinal-cord injury, ask neurologists about enrolling in registered brain-computer interface (脑机接口) trials rather than seeking unproven commercial devices.
- Follow the U.S. ClinicalTrials.gov entries for Neuralink's VOICE and Blindsight studies to track verified milestones.
- Treat any "restores sight/hearing" marketing from non-clinical vendors as unverified; check for FDA/NMPA (国家药监局) clearance status.
- For investors, distinguish device-makers with cleared or trial-backed products from those selling concepts.