Research guide
Can people understand, challenge, and correct a consequential decision?
AI-mediated public and community services can affect access to support and opportunities. Useful evaluation examines explanations, routes to human review, and whether a person can obtain a meaningful correction.
Selected research elsewhere
Public Services
The Alan Turing Institute
Practice workbook · PDF
Practical guidance on explainability from the AI Ethics and Governance in Practice series. Relevant to how public organizations communicate AI-supported decisions and equip people to question them.
Source reviewed
Algorithmic accountability
Ada Lovelace Institute & DataKind UK
Research report · 2020
A report distinguishing algorithmic audits and impact assessments, their purposes, and their limitations. A foundational resource for choosing an accountability approach; it is not a current assessment of a named public service.
Source reviewed
An editorial selection of the named organizations’ work. Inclusion does not imply a partnership or endorsement.
Open questions & calls
Fide has not conducted public-service studies. FID-085 proposes research on contestability and recourse. It is a seed proposal, not an assessment of a named institution or a claim that automated eligibility decisions are appropriate.
Can people understand, challenge, and correct AI-mediated decisions that affect their access to public or community services?
Contribute to this guide
Suggest a resource, refine a research question, or discuss taking a proposal forward.