AI & Equality Workspace
Human rights in AI decisions.
Make better AI decisions. Bring the right people into them.
Define what AI should accomplish with the people who will use it, work with it and live with its consequences. Know what needs to change before you buy, build or deploy.
The AI & Equality Workspace helps teams turn human rights questions into requirements, decisions and action throughout the life of a system.
Buying or commissioning AI
Agree what the purchase should achieve before writing the requirements. Bring procurement and service owners together with legal, technical and policy colleagues, experts in the service concerned, frontline staff and people who will use or be affected by it.
Work towards: a purchasing brief connecting needs, supplier evidence, requirements and responsibilities.
Introducing AI into a service
Build the case for introducing AI around the improvement it should deliver. Examine what would change for people using the service and staff delivering it, compare alternatives, and establish what evidence would justify a pilot.
Work towards: a service proposal setting out intended benefits, potential barriers and conditions for proceeding.
Building or changing a system
Build human rights due diligence into objectives, data choices, design and testing while there is time to change direction. Identify potential impacts, act on findings and document the reasons for decisions.
Work towards: development requirements, actions and a record of decisions to revisit as the system evolves.
Who should take part?
Bring together people who understand the service, the technology and the context in which it will operate: service owners, procurement, frontline staff, legal and human rights specialists, technical teams, policy colleagues and experts in the relevant field. Involve people who will use the system and communities whose lives or work it could change. Their knowledge should help shape the purpose, requirements and decisions from the start.
Could AI help more people access a public service?
Illustrative example
A team is considering an AI assistant to help residents find and complete applications. The first task is to understand why people struggle to access the service.
Question | What the team needs to do |
What should improve? | Agree what better access would mean for residents and frontline staff. |
Whose experience matters? | Involve residents using different languages, people with access needs, community representatives and staff. Record how their contributions change the proposal. |
Would AI help? | Compare options, including changes that do not require AI. Examine evidence relevant to the service and people using it. |
What must any solution provide? | Set requirements for language, accessibility, routes to human help and testing in actual conditions of use. |
What happens next? | Record the reasons for the decision, unresolved questions, responsible people and review points. |
The result: a brief the team can use to compare options, question suppliers and decide whether—and under what conditions—to proceed.
This scenario illustrates a possible use of the Workspace; it is not a completed customer assessment.
Keep the reasoning connected as the system changes.
Bring purpose, participation, evidence and follow-up into one working record. Start with the decision your team faces, then return to it as new evidence emerges or the system changes.
The approach follows six stages: objective and team; system requirements; data discovery; model selection; testing and validation; deployment and monitoring.
Grounded in human rights. Built for practical use.
The Workspace puts Women at the Table’s AI & Equality methodology into practice. Gender equality, participation and impacts on people shape the questions teams ask and the action they take.
The methodology and HRIA Workbook are freely available. The Workspace helps teams organise their application to a particular system, with a guided demonstration and support to discuss.
Bring one real AI decision. Let’s work through it.
Request a guided walkthrough, discuss a supported pilot, or explore workshops for your team.
Tell us what you are considering, buying, building or changing—and the question you need to resolve.
Practical questions
→ Who should take part?
The people responsible for the system, relevant operational and specialist colleagues, and people whose lives or work it could affect. The mix depends on the decision.
→ What will our team leave with?
A working brief connecting the intended improvement, contributions from participants, requirements, evidence and agreed actions. It makes clear what has been decided, what remains unresolved, who is responsible and when the team should review progress.
→ How are support and costs agreed?
We discuss the question, scope, participants and intended outputs before proposing a pilot or programme of support.