---
id: CVE-2026-12797
aliases:
  - GHSA-p897-vf7j-f5h8
title: >-
  BerriAI litellm has Security Feature Bypass in BannedKeywords and
  AzureContentSafety Guardrails via call_type Mismatch on Async Endpoints
summary: >-
  BerriAI litellm has Security Feature Bypass in BannedKeywords and
  AzureContentSafety Guardrails via call_type Mismatch on Async Endpoints
severity: medium
cvss: 6.3
cvssVector: 'CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:L'
vendor: litellm
product: litellm
ecosystem: pip
affected:
  - litellm <= 1.82.5
published: '2026-06-21'
updated: '2026-09-14'
sourceUpdated: '2026-09-14T18:27:50.648286205Z'
source: OSV
sourceUrl: 'https://osv.dev/vulnerability/GHSA-p897-vf7j-f5h8'
references:
  - url: 'https://nvd.nist.gov/vuln/detail/CVE-2026-12797'
  - url: 'https://gist.github.com/YLChen-007/078179224f07cc4e39e4f141a18c817a'
  - url: 'https://github.com/BerriAI/litellm'
  - url: 'https://vuldb.com/cve/CVE-2026-12797'
  - url: 'https://vuldb.com/submit/811288'
  - url: 'https://vuldb.com/vuln/372559'
  - url: 'https://vuldb.com/vuln/372559/cti'
  - url: 'https://github.com/advisories/GHSA-p897-vf7j-f5h8'
tags:
  - osv
  - pip
  - ghsa
epss: 0.004
epssPercentile: 0.31457
cwe:
  - CWE-285
ingestedAt: '2026-09-14T18:12:17.165Z'
---

## Overview

A security flaw has been discovered in BerriAI litellm up to 1.82.5. Affected is the function async_pre_call_hook of the file enterprise/enterprise_hooks/banned_keywords.py of the component Completions Interface. The manipulation of the argument prompt results in incorrect authorization. The attack may be performed from remote. The exploit has been released to the public and may be used for attacks. The vendor was contacted early about this disclosure.

## Affected packages

- `litellm <= 1.82.5`

## Remediation

Refer to the advisory for the patched release.