PRIMA Eye Implant: The Pinhead-Sized Chip That Just Got EU Approval to Restore Vision

TL;DR

• Science Corporation's PRIMA system — a 2mm × 2mm subretinal photovoltaic chip paired with camera-equipped glasses — received CE mark approval on 22 July 2026, clearing it for commercial sale across 30 European countries.

• It is the first brain-computer interface (BCI) device ever approved to restore form vision — the ability to read letters, numbers, and words — in patients blinded by geographic atrophy, the advanced stage of dry age-related macular degeneration (AMD).

• The first commercial implant is expected in Germany within weeks. Country-specific reimbursement applications are already underway.

• In the US, PRIMA holds FDA Breakthrough Device and Humanitarian Use Device designations but has not yet received marketing approval.

• This is not a cure. It does not restore natural vision. It is a prosthetic — and that distinction matters for every patient, clinician, and payer who will encounter it.

What Happened

On 22 July 2026, European regulators granted a CE mark to the PRIMA retinal implant system, developed by Science Corporation — the neural engineering company founded by Max Hodak, a co-founder of Neuralink. The notified body, Germany's DEKRA, authorised commercial availability across all 30 EU and EEA member states.

The approval covers patients with geographic atrophy (GA) caused by age-related macular degeneration — the leading cause of irreversible blindness in people over 60, affecting more than 5 million people globally. GA is the advanced form of dry AMD, in which photoreceptor cells in the central retina progressively die, erasing the high-acuity vision needed for reading, recognising faces, and navigating the world.

PRIMA is not a drug. It is not gene therapy. It is a prosthetic: a 2mm × 2mm wireless photovoltaic chip, thinner than a human hair, implanted beneath the retina in an outpatient procedure lasting roughly one hour. The patient wears a pair of specialised glasses containing a forward-facing camera and an eye-facing near-infrared projector. The glasses capture the visual scene, project it as patterned infrared light onto the implant, and the implant converts that light into electrical pulses that stimulate the retina's surviving bipolar cells — bypassing the dead photoreceptors entirely.

The glasses include a zoom-in feature that lets patients magnify text. The system is entirely wireless; there are no cables, no external battery packs, no percutaneous connectors.

The pivotal clinical trial, published in the New England Journal of Medicine, enrolled 38 patients across 17 clinical sites in five countries. The primary endpoint was met: patients who had lost central vision were able to read letters and words again. Critically, the implant did not degrade patients' remaining natural peripheral vision — the device is placed beneath the atrophic macula and leaves the surrounding healthy retina untouched.

Science Corporation announced the European commercial launch on the same day as the approval. The first commercial implantation is expected "soon" in Germany, where several clinical trial sites are already operational. Reimbursement applications are in progress across multiple European countries.

What It Actually Means

This is a milestone with three layers, and most coverage has only reached the first.

Layer one: the obvious. A medical device just crossed from clinical trial to commercial reality, and it restores something that was previously considered permanently lost — the ability to read. For the millions of people with end-stage dry AMD, there has been, until this week, exactly zero treatments that restore lost vision. Anti-VEGF injections treat wet AMD, not dry. Complement inhibitor drugs (like Apellis's Syfovre) slow GA progression but do not reverse damage. PRIMA is the first device that gives something back.

Layer two: the category boundary. Science Corporation is calling PRIMA a brain-computer interface — and it is, technically, since the retina is developmentally part of the central nervous system. But this is not the BCI most people picture. There is no skull incision, no electrode array penetrating cortex, no robotic insertion. The implant sits in the eye, not the brain. The "interface" is photonic, not electrical at the point of stimulation. This matters because the regulatory pathway, surgical workflow, and patient risk profile are orders of magnitude simpler than a cortical implant. PRIMA cleared CE mark through the EU Medical Device Regulation — the same framework that governs pacemakers and hip replacements — not through some novel BCI-specific pathway.

This is the quiet genius of the approach. By targeting the retina rather than the visual cortex, Science Corporation sidestepped the hardest problems in neural interfacing — biocompatibility of penetrating electrodes, signal decoding, long-term stability of cortical arrays — and landed on a system that can be implanted by a retinal surgeon in an hour. The trade-off: it only works for patients whose retinal bipolar cells are still intact, which means it is specific to photoreceptor-loss diseases like GA, not to all forms of blindness.

Layer three: the AR story nobody is telling. The PRIMA glasses are, functionally, an augmented reality system where the display is the patient's own retina. A camera captures the world. A processor converts it. A projector beams it onto a chip that stimulates neurons. This is the most commercially meaningful deployment of AR-adjacent hardware in 2026 — and it is not being sold as AR. It is being sold as a medical device. The distinction is legally necessary and strategically brilliant: medical device regulation provides a moat that consumer electronics does not.

The Signal vs. Noise Split

What this is: a genuine breakthrough in neural prosthetics — the first commercially available device that restores form vision to people who lost it. A CE mark is not a press release. It means an independent notified body reviewed the clinical data, the manufacturing process, the risk management file, and the post-market surveillance plan and concluded the benefit-risk profile is favourable. That is real.

What this is not: a cure for blindness. PRIMA does not restore natural vision. It produces artificial central vision — pixelated, low-resolution, in shades of grey — that is sufficient for reading and face recognition but is not a replacement for the biological retina. Patients retain their natural peripheral vision; the implant only covers the central macula. The system also requires the patient's retinal bipolar and ganglion cells to be intact, which means it is not suitable for glaucoma, diabetic retinopathy, or optic nerve damage.

What we don't know yet: long-term durability. The pivotal trial followed patients for 12 months. How the implant performs at 5 years, 10 years — whether the stimulation remains stable, whether the retinal cells adapt or fatigue, whether the device itself degrades — is unknown. This is normal for a first-generation implant. It is also the single most important question for patients weighing the decision to undergo surgery.

Stakeholder Landscape

Patients with advanced dry AMD (directly affected): roughly 1 million people in Europe have geographic atrophy. For those who have lost the ability to read, PRIMA offers something that did not exist last week. The procedure is outpatient, the risk profile appears manageable, and the alternative is nothing. But the vision is prosthetic, not natural — managing expectations will be the hardest part of every patient conversation.

Retinal surgeons (directly affected): a new procedure enters the repertoire. Subretinal implantation is not routine — it requires vitrectomy and precise placement beneath the retina. Training programmes and proctoring will determine how quickly this scales beyond the clinical trial sites.

Ophthalmologists and optometrists (second-order): referral pathways need to be built. Which GA patients are candidates? Who does the screening? How is candidacy determined — retinal imaging, electrophysiology, both? These questions are not yet standardised.

Payers and health systems (second-order): the device cost has not been publicly disclosed. Reimbursement applications are in progress but not yet approved in any country. The cost-effectiveness case will depend on durability — a device that lasts 5 years is a different economic proposition from one that lasts 20. Germany's early adoption suggests the case is plausible, but the negotiations are happening now.

Neuralink, Synchron, Blackrock Neurotech (competitive context): PRIMA is the first BCI to reach commercial sale for restoring a sensory function. The cortical BCI companies are targeting motor function, communication, and — eventually — vision through cortical stimulation. PRIMA's retinal approach is less ambitious and more immediately achievable. It validates the broader BCI thesis without carrying the same surgical or regulatory burden.

Max Hodak and Science Corporation (beneficiaries of the narrative): Hodak's departure from Neuralink in 2021 was widely read as a split over pace and approach. PRIMA's CE mark — achieved while Neuralink's first product (Blindsight, a cortical visual prosthesis) remains in early feasibility — is a data point in that argument. Whether the retinal and cortical approaches are competitors or complements is an open question. They target different patient populations and different failure modes of the visual system.

Cross-Layer Implications

The regulatory layer. PRIMA cleared CE mark under the EU MDR, not through any novel BCI-specific framework. This is significant because it means the existing medical device regulation is adequate — at least in Europe — for a neural interface device. The FDA pathway is different: PRIMA holds Breakthrough Device and HUD designations, which provide expedited review but also require a different evidence package. How the FDA handles PRIMA will set precedent for every BCI device that follows.

The talent layer. A commercially available retinal implant creates demand for surgeons trained in subretinal implantation, for electrophysiologists who can assess candidacy, for rehabilitation specialists who can train patients to use the system. These roles barely exist today. The first-mover advantage in building this workforce goes to the European centres that participated in the clinical trials.

The perception layer. The phrase "brain-computer interface" carries cultural baggage — cyborgs, mind control, Elon Musk. PRIMA reframes BCI as a medical prosthetic for a common disease of ageing. That reframe matters for public acceptance, for investor appetite, and for regulatory goodwill. It is harder to fear a device that helps your grandmother read again.

The AR layer. The PRIMA glasses are a functional augmented reality system that has passed medical device regulation. The hardware — camera, processor, near-infrared projector — is not fundamentally different from consumer AR glasses. The regulatory strategy of pursuing medical indication first, consumer application later (or never) is a playbook that other hardware companies will study.

What This Means for You

If you or a family member has advanced dry AMD with geographic atrophy: PRIMA is not available everywhere. The first commercial implants will be in Germany, at centres that participated in the clinical trial. Country-by-country rollout across Europe will follow as reimbursement is secured. In the US, the device is not yet approved — the FDA pathway is active but the timeline is uncertain. Ask your retinal specialist whether you might be a candidate and whether a referral to a European trial centre is feasible. The key candidacy question is whether your retinal bipolar and ganglion cells are intact — this can be assessed with optical coherence tomography (OCT) and electrophysiology.

If you are a clinician: the patient conversation is about expectations. PRIMA restores the ability to read letters and recognise faces. It does not restore natural vision. It does not restore colour perception. It does not restore peripheral vision (which GA patients typically retain anyway). The zoom-in feature helps with reading but requires the patient to learn a new way of interacting with text. The surgical risk — vitrectomy, subretinal placement — is real but appears manageable in the trial population. Candidacy screening protocols are not yet standardised; expect guidance from the clinical trial centres and professional societies in the coming months.

If you work in health policy or health system planning: the reimbursement question is the one to watch. The device cost has not been disclosed. The durability data is limited to 12 months. The patient population is large — roughly 1 million Europeans have GA — but the subset who are surgical candidates and willing to undergo the procedure is unknown. Cost-effectiveness modelling will depend heavily on assumptions about device longevity, reoperation rates, and quality-of-life benefit. Germany's early adoption provides a natural experiment; watch what the G-BA (Germany's Federal Joint Committee) does with reimbursement.

For everyone else: this is a story about what "brain-computer interface" actually means in 2026. It is not a chip in your skull. It is a 2mm photovoltaic square in your eye, paired with glasses you already know how to wear, restoring something you thought was gone forever. That is worth noticing.

Uncertainty Ledger

• Long-term durability. 12-month data is encouraging. 5-year and 10-year data does not exist. The implant is designed to be stable, but the retina is a dynamic biological environment and chronic electrical stimulation has unknowns.

• Real-world visual acuity outcomes. Clinical trial patients are selected, monitored closely, and supported by expert teams. How the system performs in routine clinical practice — with variable surgical skill, variable patient adherence, variable rehabilitation support — is unknown.

• Device cost and reimbursement. Science Corporation has not disclosed pricing. Reimbursement negotiations are ongoing. The economic case will determine how widely this scales.

• FDA timeline. Breakthrough Device designation accelerates review but does not guarantee approval. The FDA may require additional clinical data beyond what was submitted for CE mark. No public timeline exists.

• Manufacturing scalability. The PRIMA chip is fabricated using semiconductor processes. Scaling from clinical trial volumes to commercial volumes — potentially tens of thousands of units — is a non-trivial manufacturing challenge that Science Corporation has not publicly addressed.

• Competing approaches. Several companies are developing cortical visual prostheses (Neuralink's Blindsight, Second Sight's Orion). Gene therapies for dry AMD are in clinical trials. Optogenetic approaches are in development. PRIMA is first to market, but the therapeutic landscape will look different in 5 years.

Bottom Line

A 2mm photovoltaic chip implanted beneath the retina just became the first commercially available device that lets people blinded by macular degeneration read again. It is not a cure — it is a prosthetic, with all the limitations that word implies. But for patients who have been told for years that nothing can be done, "prosthetic" is a word that changes everything. The first implants happen in Germany within weeks. The rest of the world — patients, clinicians, payers, regulators — is now playing catch-up.

Sources

• Science Corporation press release, "European Commercial Launch of PRIMA," 22 July 2026. [Tier 2 — company primary source]

• New England Journal of Medicine, PRIMA pivotal trial results (referenced in multiple outlets). [Tier 1 — peer-reviewed]

• Ophthalmology Times Europe, "PRIMA, first brain-computer interface, commercially available in Europe," 24 July 2026. [Tier 2 — specialist trade]

• TechCrunch, "Science Corp wins EU approval for implant that restores central vision," 22 July 2026. [Tier 2 — technology press]

• STAT News, "Retinal implant to restore some vision loss has sights set on U.S.," 23 July 2026. [Tier 2 — health/medicine press]

• Insight News (Australia), "US firm launches 'milestone' implant to restore vision in advanced AMD," 25 July 2026. [Tier 2 — specialist trade]

• Africa Insights, "Science Corporation's PRIMA chip wins EU green light to restore vision," 24 July 2026. [Tier 3 — regional business press]

• Eyewire+, "Science Corporation Launches PRIMA in Europe Following CE Mark," 23 July 2026. [Tier 2 — ophthalmology trade]

• Silicon Republic, "Science Corp's PRIMA retinal implant gets CE mark," 23 July 2026. [Tier 2 — technology press]