# Study Shows Document Indexes Can Be Recreated From Vectors

> Researchers find that multi‑vector visual document indexes leak enough information to reconstruct pages, exposing sensitive text.

Oossa · 2026-10-08 · https://oossa.com/en/study-shows-document-indexes-can-be-recreated-from-vectors

A team led by Zhuchenyang Liu published a paper on Oct 8, 2026 showing that the vector databases used by visual document search systems can be inverted back into readable pages. The attack recovers about 47% of the words and ranks the original page first in 98.4% of queries on the ViDoRe v3 benchmark. Simple defenses like token pooling or shuffling reduce word recovery to roughly 8%, but a model that re‑orders shuffled vectors can still rank the correct page 93.5% of the time. The findings apply to another retriever as well, with source pages ranked first 70.2% of the time.

## The facts

- Paper published on 2026-10-08
- Inversion recovers 47% of words and ranks source page first 98.4% of the time

## Why it matters

Organizations storing document indexes in third‑party vector databases may need stronger protections to keep confidential text safe.

## Sources & references

1. [Inverting Multi-Vector Visual Document Indices](https://arxiv.org/abs/2610.09920) – arXiv cs.IR, 2026-10-08

Last updated: 2026-10-08