The UK's CDEI Review into Bias in Algorithmic Decision-Making (2020)

In 2019 the UK government commissioned the Centre for Data Ethics and Innovation (CDEI), an advisory body then sitting within the Department for Digital, Culture, Media and Sport, to review the use of algorithms in significant decisions about individuals. It published its final report on 27 November 2020, drawing on detailed case studies across four sectors: financial services, local government, policing and recruitment.
In recruitment specifically, the review found that automated screening and matching tools trained on a firm's historic hiring data could systematically dismiss female candidates, because the training data reflected decades of past hiring decisions skewed toward men — a textbook case of an algorithm inheriting the biases of its inputs rather than removing them.
To address this the CDEI recommended a mandatory transparency obligation on public-sector bodies using algorithms in significant decisions, and called for government guidance clarifying how the Equality Act 2010's protections against sex discrimination apply to algorithmic systems, alongside better guidance on collecting the data needed to measure bias in the first place. Independent commentators, including the Ada Lovelace Institute, pushed back on parts of the review, arguing its transparency proposals needed statutory teeth rather than voluntary guidance to be effective.
The review itself did not mandate any of its recommendations, and the CDEI was later folded into the Responsible Technology Adoption Unit in 2023; no independent post-implementation evaluation has measured whether recruitment-algorithm gender bias actually fell as a result, so this is best read as an evidenced diagnosis and policy-design template rather than a proven remedy.

Read the full analysis: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/957259/Review_into_bias_in_algorithmic_decision-making.pdf