CVE-2025-3044
mediumpriority · track · 0EPSS 0.003 · 25 %ileEURUUSEPSS · scored 2026-10-27
A vulnerability in the ArxivReader class of the run-llama/llama_index repository, versions up to v0.12.22.post1, 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 version 0.12.28.
cvss
Technical severity, 0–10. One row per scoring source.
base score 0–10 · higher = more severe · hover a vector to decode
ssvc
A stakeholder decision — act, attend, or track — from exploitation and impact.
weaknesses
The CWE weakness classes this flaw belongs to.
affected products (5)
Products and version ranges known to be affected.
advisories (3)
Vendor and ecosystem advisories describing this.
- GHSA-p7j4-jwjf-5x9wLlamaIndex vulnerability in ArxivReader class can cause MD5 hash collisions2025-07-07ghsa · osvUS
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).
- PYSEC-2025-2472025-07-07vendor · osvUS
A vulnerability in the ArxivReader class of the run-llama/llama_index repository, versions up to v0.12.22.post1, 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 version 0.12.28.
- PYSEC-2026-1569LlamaIndex vulnerability in ArxivReader class can cause MD5 hash collisions2026-07-07vendor · osvUS
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).
also known as (9)
Other names and identifiers for this entity.
sources
Which sources assert this record, and when.
references (8)
External citations — patches, advisories, analysis.
- https://access.redhat.com/security/cve/cve-2025-3044cnnvd
- https://github.com/run-llama/llama_index/commit/0008041e8dde8e519621388e5d6f558bde6ef42ecnnvd
- https://github.com/run-llama/llama_index/releasesPatchcnnvd
- https://huntr.com/bounties/80182c3a-876f-422f-8bac-38267e0345d6cnnvd
- https://github.com/run-llama/llama_index/commit/0008041e8dde8e519621388e5d6f558bde6ef42eenisa_euvd
- https://huntr.com/bounties/80182c3a-876f-422f-8bac-38267e0345d6enisa_euvd
- https://github.com/run-llama/llama_index/commit/0008041e8dde8e519621388e5d6f558bde6ef42ePatchnvd
- https://huntr.com/bounties/80182c3a-876f-422f-8bac-38267e0345d6ExploitThird Party Advisorynvd
Attackers weaponise vulnerabilities like this against internet-exposed services. See what is exposed on your perimeter with a free passive scan.
Free scan