Creating Human Wellbeing through AI.
We bring together communities, researchers, technologists and partners to make AI more accessible, responsible and beneficial to people.
Progress should be measured by human outcomes, not model capability alone.
AI is advancing quickly. We believe its progress should be judged not only by how capable a model is, but by whether it improves access, dignity, participation, learning and everyday life for real people. Octopus Foundation exists to help turn that principle into practical public benefit.
Read our missionSix areas where AI can widen participation
Our work is organised around a small number of areas where thoughtful technology and engaged communities can make a durable difference.
Accessibility
Designing so that communication, information and services reach everyone — including Deaf, disabled and multilingual communities.
AI for Good
Applying AI where it demonstrably improves access, learning, care and community capability — measured by human outcomes.
Community
Building with communities as partners: shared decisions, shared benefit and genuine participation.
Education
Practical AI literacy, responsible-AI learning and accessibility education for people at every level.
Open Research
Open, evidence-led research and evaluation — sharing what we learn where it is safe and useful to do so.
Responsible Innovation
AI proposes, humans decide. Privacy, dignity and human oversight are built in from the start.
Auslan Accessibility Initiative
A community and research initiative exploring how AI might support Auslan — Australian Sign Language — recognition, accessible communication and on-device approaches, developed with and for Deaf and Auslan communities.
Helping people put responsible AI to work
Over time we aim to help eligible founders, community organisations and social-impact projects access technical guidance, community support and responsible deployment practices — so good ideas are not held back by a lack of know-how.
Principles we hold ourselves to
Human benefit before hype
We start from the outcome for people, not from the technology or the headline.
Accessibility by design
Access is a requirement from the first line of a plan, not a feature added at the end.
Communities are participants, not test subjects
The people affected help shape the work, share in its benefit and can say no.
Open where possible, protected where necessary
We share methods, data and results openly when it is safe — and protect people when it is not.
AI proposes, humans decide
Automated systems can suggest; accountable people make the decisions that affect lives.
Evidence over claims
We describe what we have actually shown, name our limitations and avoid overstatement.
Privacy and dignity by default
We collect as little as possible, protect what we hold and treat people with respect.
There is a place for you here
Communities are participants, not test subjects. Whether you bring lived experience, research, teaching or technical skill, there is a way to take part.
Community member
Join conversations, working groups and events, and help shape priorities.
Volunteer
Contribute your time and skills to accessible, community-led projects.
Research collaborator
Partner on open research, evaluation and responsible methods.
Accessibility partner
Bring lived experience and expertise to make our work genuinely accessible.
Technology partner
Offer tools, infrastructure or engineering support for public-benefit work.
Mentor
Guide founders and community organisations putting responsible AI to work.
Social-impact founder
Explore support for building responsible, accessible products for public benefit.
Follow the work as it develops
We share updates as initiatives progress. Get in touch to follow along, collaborate or ask a question.