C003 — Assessment¶


Research	R0031 — Plural Voice Claims (Blind)
Run	2026-03-29
Claim	C003

BLUF¶

Confirmed. The KPMG/University of Melbourne press release explicitly states that 57% of employees hide their AI use. The figure is widely reported across secondary sources (TechTimes, FullStackHR, Yahoo News, HR Grapevine, WinBuzzer) all citing the same KPMG/Melbourne study.

Probability¶

Rating: Almost certain (95-99%)

Confidence in assessment: High

Confidence rationale: The statistic is stated verbatim in the KPMG International press release and confirmed by multiple secondary sources all citing the same study. Minor nuance: the claim says "48,000 workers" but the study surveyed 48,340 people (general population), of whom the 57% applies to the employed subset.

Reasoning Chain¶

FACT: KPMG International press release states "over half (57%) of employees say they hide their use of AI and present AI-generated work as their own." [SRC01-E01, High reliability, High relevance]
FACT: Multiple secondary sources (TechTimes, Yahoo News, HR Grapevine) independently report the same 57% figure, all attributing it to the KPMG/Melbourne study. [SRC02-E01, High reliability, High relevance]
JUDGMENT: The claim is accurate. The only minor nuance is that the study surveyed 48,340 people total (not 48,000 workers specifically), but the 57% figure applies to the employed subset of that sample.

Evidence Base Summary¶

Source	Description	Reliability	Relevance	Key Finding
SRC01	KPMG press release	High	High	"57% of employees hide their use of AI"
SRC02	KPMG global report page	High	High	Confirms 57% with additional context on governance gaps

Collection Synthesis¶

Dimension	Assessment
Evidence quality	Robust — primary source confirmation
Source agreement	High — all sources report the same figure
Source independence	Derived — secondary sources all cite the same primary study
Outliers	None

Detail¶

This is a straightforward statistical verification. The 57% figure is consistently reported across all sources.

Gaps¶

Missing Evidence	Impact on Assessment
Full methodology of the employed subset analysis	Low — the headline figure is confirmed

Researcher Bias Check¶

Declared biases: The researcher's plural voice advocacy makes this statistic particularly compelling for the article's narrative — workers hiding AI use supports the argument for open attribution.

Influence assessment: Low risk for accuracy verification. The statistic is factual. The interpretive risk is in how prominently it is used.

Cross-References¶

Entity	ID	File
Hypotheses	H1, H2, H3	`hypotheses/`
Sources	SRC01, SRC02	`sources/`
ACH Matrix	—	ach-matrix.md
Self-Audit	—	self-audit.md