aceModerate evidence
Published recordCrimeProblem Space
More
Artificial intelligence in criminal investigation: effects of large language model consultation on confirmation bias in a simulated investigation game
- Truvace record
- TRV-2026-2497raw data (JSON)
- Version
- v1Published
- First published
- Evidence type
- Randomized controlled trial
- Sources
- 1 sourcePeer-reviewed
- sha256
- f2b02e0ee697…check it
Objectives Confirmation bias can distort information search and evidence interpretation in criminal investigations. We tested whether consulting a large language model (LLM) could reduce this bias in group decisions.Methods In total, 108 participants formed 36 three-person teams randomly assigned to LLM consultation, theory-based information and debiasing strategies, or unaided control. Teams investigated homicide cases adapted from real cases in a web-based game. Outcomes were investigative cues opened, deliber…
The problem
In a randomized study of 36 three-person teams in a simulated homicide game, consulting a large language model did not clearly improve accuracy in assigning culpability or cut deliberation time versus unaided controls.
What this doesn’t fix
The accuracy interaction appeared only in weighted analyses and was nonsignificant in unweighted analyses, so the authors say it should be interpreted cautiously; no clear advantage of the LLM over theory-based support was established.
Sources
1 on file- Psychology, Crime & Lawdoi:10.1080/1068316x.2026.2739982
Cite this record
v1Truvace Impact Record TRV-2026-2497, v1: “Artificial intelligence in criminal investigation: effects of large language model consultation on confirmation bias in a simulated investigation game.” Truvace, 2026-11-12. /record/TRV-2026-2497 (accessed on the day you copy it). sha256 f2b02e0ee6972ce3…
Change history
1 versionEvery change made to this record since it was first published, and why. Nothing has changed since it went up.
Certified into the record
Check this record hasn’t been altered
How to check it yourself, without trusting this page
Download any version from /api/record/TRV-2026-2497 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.