Artificially Confident Research
Research, not just reaction.
Original, source-linked studies into consequential AI systems: what can be established, what remains unclear and what organisations should do next.
A bounded question
Each study starts with a question that public evidence can genuinely answer. We distinguish what was found from what is inferred.
A reproducible record
Methods, criteria, source links and downloadable datasets are published alongside the analysis wherever the evidence permits.
A correction route
Important limitations are visible. New evidence and material corrections are recorded rather than quietly replacing the original research.
Published work
Latest research
-
AI in the Investigation Room: The Tools Are Arriving Before the Operating Model Is Complete
Original research into eight UK policing AI projects finds clear investigative ambition, but much thinner public evidence on training, disclosure and challenge—and makes the case for a UK…
-
UK Police AI Transparency Index: What 10 Forces Disclose
Original research scores public AI disclosure across 10 UK police forces, with a transparent method, source trail and downloadable evidence workbook.