CVE-2026-59822: BerriAI LiteLLM Improper Authentication
BerriAI LiteLLM · listed in the CISA catalog of actively exploited vulnerabilities since 2026-09-02.
- CVSS v4.0 (CVE record)
- 8.8
- high
- EPSS (exploit probability, 30 days)
- 0.8%
- higher than 58.6% of all CVEs
- Added to CISA KEV
- 2026-09-02
- US federal due date 2026-09-16
- Automatable (CISA SSVC)
- yes
- technical impact: partial
CVE-2026-59822 explained
CVE-2026-59822 is an improper authentication flaw in BerriAI LiteLLM, a proxy server (AI Gateway) for calling LLM APIs. Before 1.84.0, the MCP Streamable HTTP endpoint could accept requests without a valid LiteLLM key, which lets unauthenticated attackers reach MCP tooling. CISA lists it as known to be exploited, so affected deployments need prompt attention.
Impact
An unauthenticated attacker can send a fabricated Authorization header that triggers an OAuth2 passthrough fallback. Failed key validation is then replaced with an empty authentication object, so requests reach MCP tooling without a valid LiteLLM key. CISA rates the flaw as automatable, meaning attackers can exploit it at scale, with partial technical impact. CISA also lists it as known to be exploited in attacks. Whether it is used in ransomware campaigns is unknown.
Who is affected
The CVE record lists BerriAI litellm versions < 1.84.0 as affected. The record's structured version data names no fixed version. However, the description states that the issue is fixed in version 1.84.0. Confirm the fixed release in the vendor advisory before you upgrade.
Written with AI from the data on this page and checked against it automatically.
Official description
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, LiteLLM's MCP Streamable HTTP endpoint allowed an unauthenticated attacker to use a fabricated Authorization header to trigger an OAuth2 passthrough fallback path that replaced failed LiteLLM key validation with an empty UserAPIKeyAuth() object, allowing requests to reach MCP tooling without a valid LiteLLM key.
This issue is fixed in version 1.84.0.
CISA required action: Apply mitigations in accordance with vendor instructions, ensuring compliance with CISA’s BOD 26-04 Prioritizing Security Updates Based on Risk guidance and CISA’s “Forensics Triage Requirements”. Follow applicable BOD 26-04 guidance for cloud services or discontinue use of the product if mitigations are unavailable. Stakeholders are responsible for evaluating each asset's internet exposure and ensuring adherence to BOD 26-04 patching guidelines.
CISA notes
Affected and fixed versions
From the CVE record published by the vendor or CVE numbering authority.
- BerriAI litellmAffectedFixed in
- < 1.84.0
not stated
| Product | Affected | Fixed in |
|---|---|---|
| BerriAI litellm |
| not stated |
How to fix it
- The CVE record names no fixed version – check the vendor advisory and CISA's required action for updates and mitigations.
- Vendor advisory: github.com/BerriAI/litellm/security/advisories/GHSA-7488-…
- CISA's deadline for US federal agencies: 2026-09-16 (passed) – a useful benchmark for everyone else, too.
What to do
- Upgrade BerriAI LiteLLM to version
1.84.0or later, which the description names as the fixed version, after checking the vendor advisory. - Apply mitigations according to the vendor's instructions, as CISA's required action states.
- Restrict network access to the LiteLLM proxy and its MCP Streamable HTTP endpoint, and keep it off the public internet where possible.
- Review logs for MCP tool requests that carry no valid LiteLLM key, and check for signs of compromise.
- If mitigations are unavailable, follow BOD 26-04 guidance or discontinue use of the product, as CISA requires.
The structured version data names no fixed version, but the description states the issue is fixed in 1.84.0. Confirm this in the vendor advisory and upgrade. Until then, apply the vendor's mitigations, limit network exposure of the MCP endpoint, and follow CISA's required action. If no mitigations are available, discontinue use of the product.
Written with AI from the data on this page and checked against it automatically.
CISA's assessment (SSVC)
- Exploitation
- Exploited in the wild
- Automatable
- Yes – attackers can automate it at scale
- Technical impact
- Partial – limited control or information
Weakness: CWE-287 Improper Authentication, CWE-306 Missing Authentication for Critical Function
Questions about CVE-2026-59822
- Is CVE-2026-59822 being exploited?
- Yes. CISA added it to its Known Exploited Vulnerabilities catalog on 2026-09-02, with a US federal due date of 2026-09-16. Whether it is used in ransomware campaigns is unknown.
- Does an attacker need valid credentials?
- No. Before
1.84.0, an unauthenticated attacker could use a fabricated Authorization header to reach MCP tooling without a valid LiteLLM key. - Which version fixes the issue?
- The description states the issue is fixed in
1.84.0. The record's structured version data lists no fixed version, so confirm the fix in the vendor advisory.
Written with AI from the data on this page and checked against it automatically.
References
Data: CISA Known Exploited Vulnerabilities (public domain) · CVE Program records incl. CISA ADP enrichment · OSV.dev (CC BY 4.0) incl. Ubuntu and Debian security data · FIRST EPSS · endoflife.date. Updated automatically – facts last changed 2026-11-21. Always confirm with the vendor's or distribution's advisory before patching.