AI Sec Watch: A Security Intelligence Platform for AI Systems

Luu, T.J.

LowVulnerability

GHSA-rwvc-j5jr-mgvh: Vercel’s AI SDK's filetype whitelists can be bypassed when uploading files

Source: GitHub Advisory(opens in a new tab)
Identifiers: CVE-2025-48985GHSA-rwvc-j5jr-mgvh
Published: November 6, 2025
Record updated: October 10, 2026

View JSON

Affected: ai < 5.0.52, fixed in 5.0.52
ai >= 5.1.0-beta.0, < 5.1.0-beta.9, fixed in 5.1.0-beta.9
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.3%

Summary

A vulnerability in Vercel's AI SDK allowed users to bypass filetype whitelists when uploading files. The flaw is tracked as GHSA-rwvc-j5jr-mgvh and is fixed in versions 5.0.52, 5.1.0-beta.9, and 6.0.0-beta.

Mitigation

Fixed in 5.0.52, 5.1.0-beta.9, and 6.0.0-beta. All users are encouraged to upgrade.

Details

Scores and exploitation

CVSS score: 3.7(low)
Patch available: Yes
CVSS vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N
Attack vector: network
Attack complexity: high
Privileges required: none
User interaction: none
EPSS: 0.3%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on October 10, 2026.
Disclosure date: November 6, 2025

Weaknesses and techniques

Weakness (CWE): CWE-20
CWE-682

Affected

Packages: ai@< 5.0.52 (fixed: 5.0.52)
ai@>= 5.1.0-beta.0, < 5.1.0-beta.9 (fixed: 5.1.0-beta.9)

Classification

Attack type: Other
Attack sophistication: Moderate
AI component targeted: API
Impact: Integrity
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative.

Original source: https://github.com/advisories/GHSA-rwvc-j5jr-mgvh

First tracked: October 10, 2026 at 12:52 AM ET, from GitHub Advisory Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 85% (self-assessed)

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

ainpmLLM dependency since 2014-02-21 · 6 tracked dependents

Summary

Mitigation

Affected packages in the Exposure Registry

Related items