Global MedTech company Ottobock has announced the public release of the #DearAI Community Library, an open-source library of images depicting the everyday lives of people with limb loss and limb difference, built by the disability community and freely available on Hugging Face.
According to Ottobock, the library addresses one of AI’s persistent blind spots: the near-total absence of people with disabilities, and people with limb loss and limb difference in particular, from AI-generated imagery.
“AI can only imagine what it has already seen and it has never truly seen the real lives of people with limb loss or limb differences. With the #DearAI Community Library, we are changing that,” said Martin Böhm, Chief Experience Officer at Ottobock.
“This initiative is about contribution: giving AI something real to learn from, built by the people whose lives it has never accurately reflected. The community shaped every decision, and that is what makes this library different.”
As AI-generated imagery becomes more prevalent across media, marketing, and daily life, research has confirmed what the disability community has long experienced firsthand: AI systems overwhelmingly fail to reflect the reality of life with a disability.
According to the research cited by Ottobock, across 444 AI-generated images spanning 37 occupations, not a single person with a visible disability appeared. When disability did appear in AI-generated images, nearly 100% of depictions defaulted to a manual wheelchair. When people with limb loss did appear, they were typically depicted using bionic or cyborg aesthetics that bore little resemblance to real life.
The release adds that these failures are not neutral: research confirms they cause measurable harm, reinforcing perceptions of disability as tragic, incapable, and other-than-normal, and erasing the diversity of lived experience to a narrow set of stereotypes.
A Library Built By the Community, For AI
The #DearAI Community Library was built on the premise that if AI learns from the data it is given, then those most affected by misrepresentation should have a voice in shaping that data.
An international group of Ottobock prosthetic users, with upper and lower limb differences, worked within Microsoft’s Community Library Creator to define preferred representation. Drawing on Ottobock’s own image libraries as well as user-submitted images and videos from around the world, ambassadors selected and annotated content against community-defined standards for AI imagery.
The result is two datasets: the Limb Difference & Prosthetic Representation Dataset – Upper Limb (LDPR–UL), and the Limb Difference & Prosthetic Representation Dataset – Lower Limb (LDPR–LL), capturing people in everyday moments as parents, colleagues, friends, athletes, and community members.
“Representation in AI matters because it shapes how people see themselves and others,” said Neil Barnett, Microsoft’s Chief Accessibility Officer.
“We’re committed to image generation models that create authentic, respectful, and positive representations of people with disabilities. Ottobock’s community library demonstrates what’s possible when communities are empowered to define good representation for themselves rather than having it defined on their behalf. This is the kind of community-led approach that responsible AI development requires.”
The library is published on Hugging Face and is freely accessible to researchers, developers, brands, creative agencies and individual creators.
Bebe Vio, Davide Morana, Ezra Frech, Zainab Al-Eqabi, Hari Budha Magar and many other community members have contributed imagery to train AI with real perspectives.
“I spent my career proving what is possible for people with limb loss. But when I searched for images of someone like me, simply living everyday life, AI came up empty,” said Noah Elliott, three-time Paralympian and two-time gold medalist, and Ottobock ambassador.
“This library is for every person with a prosthetic who has felt invisible in the world AI is building. We built it for ourselves and now it belongs to everyone.”
“At the heart of DEAR AI is something deeply human: the wish to be seen for who you really are,” said Lucas Schneider, CCO and Founder, PRESENCE.
“We did not want to speak for anyone. We wanted to listen, learn from real experiences and create something that felt honest, while showing what technology could become when more people are part of shaping it.”
The Community Library is the foundation of #DearAI, a campaign bringing together more than 60 ambassadors and community members from across the world who have written personal letters directly to artificial intelligence, sharing lived experiences and everyday realities.
The #DearAI Community Library is available on Hugging Face.
A full list of participating ambassadors and additional campaign information is available here.
