They built a task from 76 real enterprise policy documents.
Formatting loss in document ingestion makes agents confidently wrong
Enterprise agents lose meaning encoded in highlight and colour, not bold or italic, leading to high false positive rates on flagged requirements.
Big Tech
Yikai Zhao · Saurabh Pandey · Pradeep Kumar Misra
Amazon
Research Digest··2 min read
The authors show that mainstream document converters silently discard visual cues like highlight and colour that authors use to mark obligations in policy documents.
§