3 tasks, each one witnessed by the sources that watched the job — and behind every one, a prompt you can use tonight.
You’ll split time between meetings with people with disabilities, desk research on new AI tools, and writing advice for policymakers. Expect one to three user sessions a week where you gather feedback, often using a screen-reader simulator to reproduce how someone blind might experience an app.
Afternoons often go to tracking AI news — reading 5–10 research summaries or product updates — and preparing short policy briefs recommending specific reform steps, like accessibility rules for automated decision systems.
Use informed consent: clearly state what you’ll test, what data you’ll collect, and how you’ll store it. When testing AI models, avoid exposing participants to unvalidated medical or legal advice from the system; label those outputs as machine-generated and not professional guidance.
Also limit sensitive data collection, anonymize responses, and run simulator tests first to reduce obvious glitches before inviting people with disabilities to try prototypes.
Take basic courses in ethics, disability studies, and human-computer interaction (HCI). Learn to use a screen-reader simulator (NVDA is free) and practice auditing websites for accessibility (WCAG guidelines).
Do small projects: run 3–5 accessibility tests, write short policy memos, and volunteer with disability advocacy groups to gather feedback — real experience matters more than theory.
Accessibility testers focus on fixing product bugs (color contrast, keyboard navigation) and usually do hands-on QA. Policy analysts write laws and broad reports. A technology ethics consultant sits between both: you test tech (using simulators and user sessions) and translate findings into ethical policy recommendations.
You’ll need both practical testing skills and the ability to argue for legal or organizational change, for example proposing how AI vendors must document training data to protect disability rights.
Clear, evidence-based communication: you must turn user feedback and AI impact analysis into concise, concrete recommendations. That means one-page briefs with specific asks (e.g., require screen-reader compatibility tests, require audits of AI decision-making for bias) and citations to the user sessions and simulator tests.
Policymakers respond to numbers and stories: include a short statistic (e.g., '3 of 5 participants could not complete checkout with the screen-reader simulator') plus a direct policy action to fix it.