Accessible Intelligence
Accessible Intelligence
Advancing research on AI-generated intelligence that is accessible, understandable, controllable, and beneficial for everyone.
Accessible Intelligence is a research initiative exploring how people interact with, understand, and benefit from increasingly intelligent systems. The initiative brings together Human-AI Interaction, accessibility, artificial intelligence, immersive technologies, and human-centred design to ensure that future AI systems support diverse users, contexts, and goals.
The Vision
Artificial intelligence is transforming how people obtain information, make decisions, create content, and interact with digital systems. Yet much of today's AI research focuses on improving model capability rather than improving people's ability to understand, question, supervise, and effectively use AI-generated intelligence.
Accessible Intelligence seeks to shift this perspective by ensuring that AI-generated intelligence is accessible, understandable, controllable, and beneficial for people with diverse abilities, expertise, contexts, and goals. Rather than viewing accessibility as an afterthought, the initiative places it at the centre of intelligent system design, evaluation, and deployment.
Why Accessible Intelligence?
🧠 Human Understanding
Understanding how people interpret, reason with, and make decisions using AI-generated intelligence.
♿ Accessibility
Designing AI systems that support people with diverse abilities, needs, expertise, and interaction preferences.
🤝 Human Agency
Ensuring people remain active decision-makers through meaningful supervision, intervention, and control tailored to diverse user needs.
🔍 Transparency
Developing explanations and interactions that help people with diverse needs and abilities understand and interrogate AI behaviour.
🛡 Trust
Supporting appropriate trust for diverse user needs and abilities through transparency, evaluation, calibration, and effective interaction design.
⚖ Responsible AI
Connecting accessibility, ethics, governance, and human-centred design within intelligent systems.
Research Agenda
Human-AI Interaction
Designing interactions that enable effective and accessible collaboration between people and intelligent systems.
Accessible Explanations
Developing explanations that can be understood by people with diverse cognitive abilities, backgrounds, and expertise.
Human Agency
Investigating supervision, intervention, contestability, reversibility, and meaningful human control for people with diverse needs and abilities.
Agentic AI
Understanding how increasingly autonomous AI systems can remain aligned with diverse human goals, needs and oversight.
AI & Immersive Technologies
Exploring how AI and extended reality combine to create accessible, trustworthy, and intelligent immersive experiences.
Evaluation Methods
Developing new methods, metrics, and frameworks for evaluating accessible intelligent systems.
Current Projects
Accessible Intelligence Framework
A conceptual framework defining Accessible Intelligence and its role within future Human-AI Interaction.
XR Accessibility Toolkit
Research software supporting accessibility evaluation in virtual, augmented, and mixed reality environments.
Alternative Text Research
Improving AI-generated alternative text through contextual information and human-centred evaluation.
ECAM-VI
A framework for understanding visual impairment as lived experience across physical, digital, and emerging interaction contexts.
Publications
Selected publications contributing to the Accessible Intelligence research agenda will be available through the Publications page. This section will continue to grow as the initiative develops.
Join the Initiative
Researchers
Collaborative research, interdisciplinary partnerships, and joint grant applications.
Students
PhD opportunities, MSc dissertations, internships, and research projects.
Industry
Knowledge exchange, consultancy, collaborative innovation, and industrial research partnerships.
Funding
Developing collaborative projects that advance the future of accessible, trustworthy, and human-centred AI.
Accessible Intelligence
A research initiative exploring how AI-generated intelligence can become accessible, understandable, controllable, and beneficial for everyone.