Skip to content
LiteLLMGHSA-p897-vf7j-f5h8

BerriAI litellm has Security Feature Bypass in BannedKeywords and AzureContentSafety Guardrails via call_type Mismatch on Async Endpoints

Low6.3CVE-2026-12797 · Published Jun 21, 2026 · updated Sep 14, 2026

GitHub advisory

Affected versions

PackageAffectedFixed in
litellm
PyPI
<= 1.82.5No fix yet
Details and references

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.

CVSS 3.1
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:L
Severity from
GitHub (reviewed advisory)
Weakness
CWE-285
Also known as
CVE-2026-12797

More LiteLLM advisories

All LiteLLM
DateAdvisory
Jun 21LiteLLM: Admin Key Handler Has Improper Authorization
CVE-2026-12770Low5.4no fix yet
Jun 21LiteLLM: PROXY_ADMIN database API Key Generator Has Insufficient Session Expiration
CVE-2026-12772Low6.3no fix yet
Jun 21LiteLLM: M2M JWT Handler Has Improper Authorization
CVE-2026-12771Low5.0no fix yet
Jun 21LiteLLM: MCP Proxy Has Improper Authentication
CVE-2026-12773Medium7.3fixed in 1.84.0
Jun 21LiteLLM: SSO Debug Flow Has Improper Authentication
CVE-2026-12795Medium7.3no fix yet
Jun 21BerriAI litellm has SSRF via Unvalidated spec_path URL in MCP OpenAPI Spec Loader
CVE-2026-12798Low6.3no fix yet

Critical advisories by email

Wednesdays: the week’s critical and high advisories in the AI and data stack, with the fixed versions. Only in weeks that have some.

Double opt-in. Unsubscribe any time.