Annual report 2025–26
The year from 1 April 2025 to 31 March 2026. Approved by the Board of Directors on 18 June 2026 and by the Board of Councillors on 25 June 2026, and published on 30 July 2026.
From the chair
Every year someone asks me whether the Foundation is really independent of the company my grandfather started. It is a fair question, and this report is, in a sense, our answer to it: everything we did with the company’s dividends, how well it worked, and where it did not.
It was a good year for the company, and therefore for us. Kiroshi Opticals paid ¥19.7 billion in dividends to the Foundation, 15% more than the year before, and we spent ¥24.3 billion on charitable work, the most in our history. I should be plain about where the increase came from. The company’s medical division grew modestly. Most of its growth came from its Professional Systems division, which makes optics, implants and perception software for security and defence customers. Our grants are paid for, in part, by that business. I do not think we should pretend otherwise, and I do not think our grantees would want us to.
We changed how we count. From this year, we estimate the outcomes of cataract patients who do not come back for follow-up, by tracing a sample of them at home. They did less well than the patients who returned, and our headline figure for good outcomes is now a range, 72% to 78%, rather than a single flattering number. I would rather we published the range.
Frothly’s Blackwall programme published its evaluation of SHIKAI 4, the perception model in the company’s newest implants, in November. It found that the model recognised white canes and guide dog harnesses less reliably in low light than in daylight: a failure that matters most to exactly the people our programmes serve. As a shareholder, we asked the company to fix it before release and to publish the results of retesting. It has done the first and promised the second. I am grateful to Frothly for finding it, and to the company for agreeing in 2023 to let them look.
Continuity has now helped nearly nine thousand people with SEIRAN 2 implants, a device the company stopped supporting in 2024. I visited the repair clinic in Oakland in January. A woman there showed me the old battery from her eye, which she keeps on her windowsill. I keep thinking about it.
Finally, the board has asked for the 1996 Arrangement between the Foundation and the company to be reviewed. It was written before the company sold SEIRAN, its first implant with software in it, and it says nothing about data, firmware or the devices people carry inside them. We expect to publish proposals in 2027.
Kiroshi Reiko
Chair of the Board of Directors
Tokyo, June 2026
The year in figures
- charitable spending
- ¥24.3bn
- cataract operations
- 61,240
- good outcome, including patients not followed up
- 72–78%
- people screened
- 1,482,906
- pairs of spectacles
- 74,113
- reports to the Implant Register
- 27,416
- safety signals issued
- 14
- people with discontinued implants helped
- 11,870
- people in low vision rehabilitation
- 9,320
- new grants through open calls, 44 of them for ¥5 million or more
- 463
- Shared Sight donors
- 68,402
- staff in six offices
- 214
The Executive Director’s review
Dr Okabe Sena writes: Most of what we did this year was the same work as last year, done a little better and for more people. Sight restoration funded 61,240 cataract operations, up 6%, and the follow-up rate reached 71.2%. Mieru screened nearly a million and a half people and, more importantly, raised the share of children still wearing their spectacles after three months to 58%. The Implant Register issued fourteen signals, its most ever, and its median time from first report to signal fell to nineteen days.
Three things were new. We began tracing patients who miss their follow-up, which has made our figures less flattering and more true. We opened two aftercare clinics, in Cebu and Chicago. And Continuity began buying SEIRAN 2 batteries in bulk, which halved the waiting time for a replacement.
Our research partnership with Frothly continued. KF-LV, the low-vision test set we built with Blackwall, was released in March 2025 and was used in Blackwall’s evaluation of SHIKAI 4, which is how the white cane failure was found. A test set built for blind people found a fault that mattered to blind people. That is the point of independent evaluation, and I hope other makers will ask for it.
What went wrong
- Our cataract outcomes were worse than we had been saying. Tracing patients at home showed that those who miss follow-up do worse. The true share of good outcomes is between 72% and 78%, not 78.4%.
- Spectacle wear in Tanzania and Kenya remained below half. We have changed the frames, added spares and moved delivery to the start of term. We will report next year whether it helped.
- Two independent studies found our effect smaller than we claimed. Bhattarai, Khadka and Aaltonen (2025) found that about a third of the children Mieru gave spectacles to in Nepal would have received them within two years anyway. Ochieng and Varela (2026) found that about one in six of the cataract operations we paid for in two Kenyan counties would have happened within a year through public provision. We have stopped describing every pair of spectacles and every operation as one that would not otherwise have happened. Both studies are summarised, with full references, on the research page.
- We ended one grant for fraud. A screening partner in one district reported schools it had not visited. We recovered ¥14.2 million of ¥31 million paid, and reported the matter to the police.
- The SEIRAN 2 is ageing faster than its spares. Some parts, including one connector, are no longer made by anyone. Continuity is paying for a new one to be designed and published openly; until then, some repairs will be slower.
The accounts in brief
- Income
- ¥24.68 billion, of which dividends from Kiroshi Opticals ¥19.70 billion
- Charitable spending
- ¥24.30 billion
- Endowment, 31 March 2026
- ¥684.2 billion, up from ¥634.7 billion
- Held in Kiroshi Opticals
- 66% of the endowment
- Auditors’ opinion
- Unqualified, signed 12 June 2026
Earlier annual reports
Annual reports from 2016–17 onwards are available as PDFs on request from info@kiroshi.org, in English or Japanese, and in print. Reports before 2016 are held in the Founder’s Workshop and can be read there.
- 2024–25¥22.4bnThe year SEIRAN 2 support ended. First publication of outcomes by hospital.
- 2023–24¥20.9bnKiroshi Opticals agrees to Blackwall evaluation. KF-OS 1 version 2.1.
- 2022–23¥19.6bnShared Sight opens. Dr Okabe Sena becomes Executive Director.
- 2021–22¥18.1bnThe research partnership with Frothly begins.
- 2020–21¥16.2bnContinuity founded. Mieru paused for six months during school closures.
- 2019–20¥17.4bnSEIRAN 2 discontinued; Raven Microcybernetics dissolved. The last sale of Kiroshi Opticals shares.
- 2018–19¥16.9bnThe Founder’s Workshop reopens as a museum.
- 2017–18¥15.8bnSecond aftercare clinic opens, in Osaka.
- 2016–17¥15.1bnThe Chiba Centre and the Independent Implant Register open.