All reports
Medium

LlamaIndex vulnerability in ArxivReader class can cause MD5 hash collisions

Affected versions

Version ranges need review.

Affected versions
< 0.3.1
Fixed version
0.3.1
Affected versions
<0.3.1
Fixed version
0.3.1
Affected versions
<0.12.28
Fixed version
0.12.28

Exploitation

CVE-2025-3044

Probability of exploitation in the next 30 days: 0.32%

FIRST EPSS · 2026-11-18

Description

A vulnerability in the ArxivReader class of the run-llama/llama_index repository allows for MD5 hash collisions when generating filenames for downloaded papers. This can lead to data loss as papers with identical titles but different contents may overwrite each other, preventing some papers from being processed for AI model training. The issue is resolved in llama-index-readers-papers version 0.3.1 (in llama-index 0.12.28).

Links