Threat
Joshua Judges Ruth
First reported tomclancy.info
Page published
Earliest dated coverage: 7 Nov 2026 · First observed: 7 Nov 2026 · Latest dated coverage: 7 Nov 2026
Coverage timeline
Single-source research — one report is available.
Why it matters
Document-borne prompt injection and hidden PDF content can corrupt financial data extracted by AI systems, even when the application is internal and its inputs appear legitimate.
An internal AI financial-document parser reportedly accepted manipulated earnings figures embedded in PDFs using visually hidden, off-page, or unusually sized text. The QA write-up also describes a malicious footnote impersonating a recent SEC rule change and instructing the agent to report a fixed value for every number.