The Wire
A talking classroom in Malawi is teaching blind pupils to learn on their own
Staff Nyoni built Blind Classroom after watching his blind grandfather struggle to read. Now the voice-led AI system is turning lessons into conversation at schools for the blind across Malawi.
Key points
- Al Jazeera reported on 16 November 2026 that an artificial intelligence tool built by 25-year-old Malawian innovator Staff Nyoni is helping blind and visually impaired pupils improve their school results. Al Jazeera, 16 Nov 2026
- The system, Blind Classroom, converts written lessons into interactive audio that pupils navigate with a three-button interface and voice commands. D+C, 11 Mar 2026
- Yankho Maganga, 14, a pupil at Chilanga School for the Blind in Kasungu District, says her grades were very poor before the system and that she is now among the top performers. D+C, 11 Mar 2026
- The system is deployed at Chilanga School for the Blind and St Joseph Resource Centre in Dedza; 20 systems are installed and 70 more are ready, against a target of 140 across 14 schools. NCST Malawi, 1 Jul 2025
- Malawi has more than 800,000 visually impaired people, yet only 50,964 are enrolled in schools; 25,396 male and 25,568 female. NCST Malawi, 1 Jul 2025
ONE EVENT. THREE READINGS.
The story reads as development reporting with a twist.
D+C, the German development magazine, sent a reporter to Chilanga and let the pupil and the teacher carry the story, and the Western lens lingers on the modesty of the money: €18,000 to reach 14 schools, against the billions that dominate the AI debate in Washington and Brussels.
What this reading highlights: The Gates Foundation meeting is read as a signal that the world’s largest philanthropy is watching a tool built far from Silicon Valley.
The Western question is whether a system built for Malawi’s curriculum can travel, and who pays when it does.
Sources: D+C (1)
In the East, there is, honestly, no independent reporting to quote.
Searches of CNA, The Hindu, NHK and Kyodo found no standalone account of Blind Classroom, and the only Asian carriage of the story is Rising Nepal Daily in Kathmandu, which republished the D+C report; one evidence chain, not a second one.
What this reading highlights: The Eastern lens here is the absence itself: a story about artificial intelligence serving blind children in Africa travelled through a German magazine and a Nepali republication rather than through Asia’s own newsrooms.
The one Eastern institutional footprint is diplomatic: Japan’s ambassador to Malawi hosted Nyoni at an innovation gathering in Lilongwe.
Sources: Rising Nepal Daily (1)
The story belongs to Malawi’s own record.
Al Jazeera, reporting from Doha, put the report in front of a global audience today; Malawi’s NCST registry carries the technical case; the Muuni Fund and Save the Children Malawi carry the funding; Project Malawi, reporting from inside the country, carried the Nairobi award.
What this reading highlights: The Southern lens centres the origin: a young man from Mzimba solving a problem he watched his grandfather live with, and a country with 800,000 visually impaired people and 50,964 of them in school, building its own answer instead of waiting for one to be imported.
Sources: Al Jazeera (1), National Commission for Science and Technology, Malawi (1), Project Malawi (1), Africa Press (1), Muuni Fund (1)
Sources by lens: 1 West, 1 East, 5 Global South. See every source
THE POINT OF CONVERGENCE
On Monday 16 November 2026, Al Jazeera reported that an artificial intelligence tool developed by 25-year-old Malawian innovator Staff Nyoni is helping blind and visually impaired pupils improve their school results; the system, Blind Classroom, turns written lessons into interactive audio at Chilanga School for the Blind in Kasungu District and St Joseph Resource Centre in Dedza, and is scaling from two pilot schools to 14 schools across Malawi.
WHAT REMAINS CONTESTED
On what the story proves; the Western development press reads it as a model of frugal African innovation, the Southern record reads it as a Malawian answer to a Malawian problem, and the absence of independent East Asian reporting leaves the Eastern lens an honest blank; whether a handful of strong individual cases amounts to proof of learning gains is openly contested.
The story
What did a 25-year-old from Mzimba build for pupils who cannot see the blackboard?
Al Jazeera reported on Monday 16 November 2026 that an artificial intelligence tool developed by a 25-year-old Malawian innovator is helping blind and visually impaired pupils improve their results at school. The system is called Blind Classroom, and at Chilanga School for the Blind in Kasungu District, in central Malawi, pupils who once depended on Braille books and constant help from teachers now follow lessons by listening, answering aloud and repeating as often as they need. Yankho Maganga, 14, a pupil at Chilanga, told the German development magazine D+C that until 2024 she relied heavily on Braille books and constant assistance from teachers to keep up with her lessons, and that her grades were very poor. Today she counts herself among the top performers in her class, and she credits the black computer sitting on the small wooden desk in front of her. The report is the latest chapter of a project that began in 2022, when Nyoni entered a Save the Children hackathon with a prototype; it is now being rolled out from two pilot schools to 14 schools for the blind across Malawi.
How does a lesson become a conversation a blind pupil can follow?
The system converts written lessons into interactive audio. Instead of tablets, pupils navigate through a simplified three-button interface and voice commands, teacher John Makombe told D+C: the device talks to them, reads out lessons, asks questions and lets them answer in their own voices. An inbuilt AI teacher explains and rephrases concepts, adjusts to each pupil’s pace and understanding, and gives real-time feedback. The lessons are aligned with the Malawian curriculum, and a whole curriculum can sit on a single device.
The design is deliberately simpler than the screen readers it replaces. Where JAWS needs more than 50 shortcut keys and NVDA more than 40, Blind Classroom uses a three-key combination on a custom keyboard, according to Malawi’s National Commission for Science and Technology, which lists the system in its innovation registry. The rollout is a partnership of Save the Children International, the Muuni Fund of Malawi’s communications regulator MACRA, and the Ministry of Education. It is already deployed at Chilanga and at St Joseph Resource Centre in Dedza District; 20 systems are installed and 70 more are ready, against a first-phase target of 140 systems across 14 inclusive schools. Save the Children has put MK38 million behind the project and budgeted about €18,000 for the scale-up, its innovation manager Bright Chidzumeni told D+C.
Nyoni, from Euthin Village in Mzimba District, studied ICT and software engineering at Daeyang University from 2019 to 2023. He says the idea came from watching his visually impaired grandfather struggle to read, and from a girl in his community who left school because she could not cope with the reading demands. In 2022 he answered a Save the Children call for applications, pitched at a hackathon run with partners, and won a two-year incubation at NextGen Labs to develop the system. The project has collected recognition along the way: overall winner of the 2025 Innovation Jam at the ICT Association of Malawi’s annual meeting, the Inclusive Africa Innovations Award at the seventh Inclusive Africa Conference in Nairobi in 2026, and a showcase at TiCon Africa 2026 in Zambia. On 4 November, Nyoni met a senior programme officer of the Gates Foundation to explore collaboration on inclusive education, according to his company, Access Ability Africa.
What has nobody measured, priced or promised yet?
No independent published study of learning outcomes exists yet; the evidence is individual cases and the testimony of teachers. The cost of one system has not been disclosed, so the price of reaching every blind pupil in Malawi cannot be calculated from public figures. It is not known whether the Ministry of Education will adopt the system nationally or fund it beyond the pilot, or whether the 70 systems described as ready are already in classrooms. The system is designed to be customised to local languages, but no public record says which languages it speaks in classrooms today. Nor is it known what the Gates Foundation meeting will lead to, or when every subject will be loaded; teacher John Makombe’s own caveat is that some important topics are still missing. One detail differs between accounts: D+C describes the relative whose struggle inspired Nyoni as his father; Nyoni himself, quoted elsewhere, speaks of his grandfather.
Why does a talking computer matter more than a new Braille book?
Because of the scale of the gap it addresses. Malawi has more than 800,000 visually impaired people, yet only about 50,964 are enrolled in schools; 25,396 male and 25,568 female, according to the NCST. The rest face a shortage of trained special-needs teachers, a lack of accessible materials and a reliance on printed content, and many drop out early. A Braille book is expensive, bulky and static; a screen reader is a general-purpose tool with a steep learning curve. A pupil who cannot read the lesson cannot keep up, and the lesson moves on without her.
Blind Classroom changes who controls the pace: the pupil repeats the lesson until it is understood, asks the machine to explain it differently, and answers without waiting for a teacher. That matters beyond the classroom, because the same bottleneck, reading, is what kept Nyoni’s grandfather from books and a girl in his village from school. The wider context is that artificial intelligence is usually debated as a race between great powers over chips and regulation, while here it arrives as a small computer on a wooden desk in Kasungu. Malawi’s own institutions are backing it: the NCST frames it under Vision 2063, the country’s long-term development plan, and the Ministry of Education is a partner in the rollout. Malawi’s factbook profile and its population figures show the scale of the need the system is built for.
What does Blind Classroom change about who gets to learn?
It changes the price of access, the shape of the lesson and the direction of the innovation. On price: a single device carrying a whole curriculum replaces shelves of Braille books and hours of teacher reading time, and the scale-up budget is measured in thousands of euros, not millions. On shape: the lesson stops being a text the pupil cannot see and becomes a conversation the pupil controls, repeat, rephrase, answer, which is a different pedagogy from a screen reader reading a page aloud. On direction: the innovation flows from the community that needs it, through a Malawian start-up, a regulator’s fund and a ministry partnership, rather than arriving as an imported product; NCST backing under Vision 2063 gives it a state frame without state funding yet.
The test will be whether that frame holds when the pilot money runs out: 14 schools is a beginning, not a system, and the 750,000 visually impaired Malawians outside school are the measure of what remains. The recognition, the Nairobi award, the Zambia showcase, the Gates Foundation meeting, buys attention; it does not buy computers, power or subjects. Malawi’s leadership has the file now; the question is whether the state picks it up or leaves it to donors.
What happens if the classroom keeps talking?
If the 140 systems land, hundreds of pupils get an independence no Braille book offered: lessons on demand, at their own pace, without a teacher at their shoulder. If the model travels, the NCST says it can be customised to any national curriculum, other countries with the same shortage of teachers and materials could adopt a design proven in Malawi rather than imported from elsewhere. If the Gates Foundation engagement turns into funding, the 14 schools could become a national programme; if it does not, the project stays a donor-funded pilot with a hard edge at 14. And if the subjects are completed and the devices maintained, the second-order effect is a generation of blind pupils who arrive at secondary school having kept up, instead of having dropped out.
What should readers watch for next?
Watch the rollout first: whether the 70 systems described as ready reach classrooms in this school year, and whether the 14-school target is met. Watch the Ministry of Education, for any move from partnership to adoption or funding. Watch for the missing subjects to be loaded, for an independent evaluation of learning outcomes, and for what follows the Gates Foundation meeting. And watch whether the design travels; the next Blind Classroom outside Malawi would be the moment a Malawian answer becomes a pattern.
What we agree on, and what remains open
Established
- On Monday 16 November 2026, Al Jazeera reported that an artificial intelligence tool developed by 25-year-old Malawian innovator Staff Nyoni is helping blind and visually impaired pupils improve their school results.
- The system, Blind Classroom, turns written lessons into interactive audio at Chilanga School for the Blind in Kasungu District and St Joseph Resource Centre in Dedza, and is scaling from two pilot schools to 14 schools across Malawi.
Contested
- On what the story proves; the Western development press reads it as a model of frugal African innovation, the Southern record reads it as a Malawian answer to a Malawian problem, and the absence of independent East Asian reporting leaves the Eastern lens an honest blank.
- Whether a handful of strong individual cases amounts to proof of learning gains is openly contested.
Still open
- What one system costs, whether the Ministry of Education will adopt or fund the system nationally, which languages the classroom AI speaks in practice, how many of the 90 prepared systems are already in classrooms, what the Gates Foundation meeting will lead to, or when every subject will be loaded; and one detail differs between accounts, with D+C describing Nyoni’s inspiration as his father and Nyoni himself speaking of his grandfather.
Questions, answered
Blind Classroom is an AI-powered, voice-controlled learning system built in Malawi for blind and visually impaired pupils. It turns written lessons into interactive audio: the device reads a lesson aloud, asks questions and accepts spoken answers, with an inbuilt AI teacher that explains, rephrases and adapts to each pupil’s pace. Pupils navigate with a simplified three-button interface and voice commands, and a whole curriculum fits on one device.
Staff Nyoni is a 25-year-old Malawian innovator from Euthin Village in Mzimba District and the founder of Access Ability Africa, the start-up behind Blind Classroom. He studied ICT and software engineering at Daeyang University from 2019 to 2023, and says the system was inspired by watching his visually impaired grandfather struggle to read. He built the first prototype after a 2022 Save the Children hackathon.
More than 800,000, according to Malawi’s National Commission for Science and Technology; yet only 50,964 of them are enrolled in schools, 25,396 male and 25,568 female. The rest face a shortage of trained special-needs teachers, a lack of accessible learning materials and a reliance on printed content, and many drop out of school early as a result.
Blind Classroom is deployed at Chilanga School for the Blind in Kasungu District and St Joseph Resource Centre in Dedza District, the two pilot schools. Twenty systems are installed and 70 more are ready, against a first-phase target of 140 systems across 14 inclusive schools in Malawi. New deployments have also begun at resource centres in Chikwawa and Malingunde, according to Access Ability Africa.
MAGNA BUREAU