---
id: CVE-2025-3044
aliases:
  - GHSA-p7j4-jwjf-5x9w
  - PYSEC-2026-1569
title: LlamaIndex vulnerability in ArxivReader class can cause MD5 hash collisions
summary: LlamaIndex vulnerability in ArxivReader class can cause MD5 hash collisions
severity: medium
cvss: 5.3
cvssVector: 'CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N'
vendor: llama-index-readers-papers
product: llama-index-readers-papers
ecosystem: pip
affected:
  - llama-index-readers-papers < 0.3.1
patched:
  - llama-index-readers-papers 0.3.1
published: '2025-07-07'
updated: '2026-07-07'
source: OSV
sourceUrl: 'https://osv.dev/vulnerability/GHSA-p7j4-jwjf-5x9w'
references:
  - url: 'https://nvd.nist.gov/vuln/detail/CVE-2025-3044'
  - url: >-
      https://github.com/run-llama/llama_index/commit/0008041e8dde8e519621388e5d6f558bde6ef42e
  - url: >-
      https://github.com/run-llama/llama_index/commit/f69e1c0e7579228fec4cfaf716e4f951e131de77
  - url: 'https://github.com/run-llama/llama_index'
  - url: 'https://huntr.com/bounties/80182c3a-876f-422f-8bac-38267e0345d6'
tags:
  - osv
  - pip
epss: 0.00284
epssPercentile: 0.21142
ingestedAt: '2026-07-08T18:25:51.728Z'
---

## Overview

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).

## Affected packages

- `llama-index-readers-papers < 0.3.1`

## Remediation

Upgrade to a patched release:

- `llama-index-readers-papers 0.3.1`