Context
DeepTRACE introduces an eight-dimension audit framework spanning answer text, sources, and citations, and applies it to generative-search and deep-research systems. The study reports substantial unsupported-statement rates and citation accuracy ranging roughly 40–80% across evaluated systems.
What changed
The work operationalizes citation reliability beyond citation presence by connecting statement-level claims to listed sources and factual support. Deep-research configurations improved citation thoroughness and reduced overconfidence in the study but did not eliminate unsupported statements or one-sidedness.
Why it matters
AI visibility measurement can overstate source quality when it counts citations without checking whether the cited source actually supports the generated statement. Reliable measurement therefore needs citation correctness and support, not only citation frequency.
Visibility OS interpretation
Visibility OS interprets DeepTRACE as evidence for separating source retrieval, citation allocation, and evidentiary support in measurement. The study does not establish that its measured error rates apply to current versions of the named systems.
What remains uncertain
How stable are these citation-support measurements across model updates, locations, languages, query classes, and repeated runs? How much do deep-research modes improve support under matched retrieval conditions?