CVE-2026-120334 deepset-ai CVE debrief

Who should care

Defenders responsible for systems or applications using deepset-ai farm up to 0.8.0 should assess exposure to remote attacks and prioritize verification of potential vulnerabilities. Security teams and administrators of affected products should monitor for potential attacks and implement compensating controls if patches or updates are not available.

Why it matters

CVE-2026-120334 is a path traversal vulnerability in deepset-ai farm up to 0.8.0, allowing remote attackers to manipulate the vocab_filename argument. The exploit has been published, and defenders should prioritize verifying exposure, monitoring for attacks, and implementing compensating controls. Affected products are no longer supported by the maintainer.

• Verify potential exposure to remote path traversal attacks
• Monitor for potential attacks due to published exploit
• Implement compensating controls for unsupported versions
• Assess priority for replacing or discontinuing use of vulnerable versions

Technical summary

The EmbeddingTokenizer.from_pretrained function in farm/modeling/tokenization.py is vulnerable to path traversal attacks due to improper handling of the vocab_filename argument. This issue affects deepset-ai farm versions up to 0.8.0 and has been publicly disclosed with a CVSS score of 2.1. The vulnerability allows remote attackers to manipulate the vocab_filename argument, potentially leading to path traversal attacks. The exploit has been published, and although the vendor was contacted, no response was received. This vulnerability affects products that are no longer supported by the maintainer.

Defensive priority

Defenders should prioritize verifying if any systems or applications using deepset-ai farm up to 0.8.0 are still in use and assess exposure to remote attacks. Given the exploit's publication, defenders should also monitor for potential attacks and implement compensating controls if patches or updates are not available.

Recommended defensive actions

• Verify if any systems or applications using deepset-ai farm up to 0.8.0 are still in use and assess exposure to remote attacks.
• Monitor for potential attacks and implement compensating controls if patches or updates are not available.
• Consider replacing or discontinuing use of unsupported deepset-ai farm versions.
• Review the supplied official advisory or CVE record to validate affected scope, severity, and vendor guidance.
• Plan vendor-supported updates or mitigations through normal change control where exposure is confirmed.
• Check relevant monitoring, detection, and logs for exposed assets that need extra review.
• Track exceptions, retest remediated assets, and close the item only after evidence is documented.

Evidence notes

The CVE record and NVD detail page provide information on the vulnerability, including its CVSS score of 2.1 and severity of LOW. The exploit has been published, but there is no information on in-the-wild exploitation or specific victims.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.