CVE-2026-42271
BerriAI LiteLLM · AI and automation
BerriAI LiteLLM Command Injection Vulnerability
At a glance
Exploited. Added to CISA’s catalog of exploited vulnerabilities on June 8, 2026.
Affected product: BerriAI LiteLLM (category AI and automation, indicative classification). Name in CISA’s catalog: BerriAI LiteLLM.
What should I do?
Apply the patches or mitigations from BerriAI; if none are available, stop using the product. For cloud services, follow the guidance from BerriAI.
Deadline set by CISA for US federal agencies: 14 days (due June 22, 2026). It only binds those agencies, but it is a useful measure of urgency.
- The vendor’s specific advisory is listed in the NVD references: CVE-2026-42271 on the NVD website.
Original text of the required action (CISA)
Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.
Timeline
- Vulnerability published Publication date (NVD)
- Exploitation confirmed Added to CISA’s KEV catalog
- CISA deadline Remediation deadline set for US federal agencies
The facts, with their sources
- Exploitation confirmed
- Yes, on June 8, 2026 Date added to the catalog of exploited vulnerabilities. Exploitation itself may have started earlier. Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- Ransomware
- No known use to date That does not guarantee it is not being used: the information is not public. Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- Hunt for compromise
- Not requested by CISA Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- CISA required action
- Apply the patches or mitigations from BerriAI; if none are available, stop using the product. For cloud services, follow the guidance from BerriAI. Rewritten by this site from CISA’s original text. Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- CISA deadline
- June 22, 2026, 14 days Set for US federal agencies: a measure of urgency, not an obligation elsewhere. Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- Vulnerability published
- May 8, 2026 Source: National Vulnerability Database (NVD), collected September 29, 2026 at 15:00 UTC
- Product according to CISA
- BerriAI LiteLLM Filed by this site under “AI and automation” (indicative classification). Source: KEV catalog, collected September 29, 2026 at 14:57 UTC
- CVSS vector
- CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H Technical detail of the score: attack vector, complexity, impact. Source: National Vulnerability Database (NVD), collected September 29, 2026 at 15:00 UTC
Severity and activity
FIRST’s EPSS model. Since exploitation is already confirmed, this figure is mostly useful to compare activity. Collected Sep 29, 2026.
Description
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full server configuration in the request body, including the command, args, and env fields used by the stdio transport. When called with a stdio configuration, the endpoints attempted to connect, which spawned the supplied command as a subprocess on the proxy host with the privileges of the proxy process. The endpoints were gated only by a valid proxy API key, with no role check. Any authenticated user — including holders of low-privilege internal-user keys — could therefore run arbitrary commands on the host. This issue has been patched in version 1.83.7.
Official description from the NVD, collected Sep 29, 2026. CVE® description © The MITRE Corporation.
BerriAI LiteLLM contains a command injection vulnerability that could allow any authenticated user, including holders of low-privilege internal-user keys, to run arbitrary commands on the host.
Summary from CISA (KEV catalog).
CERT-FR advisories
No advisory among the latest CERT-FR publications this site follows. That does not mean there are none.
Information reused under the Open Licence 2.0 (Etalab); the date of last update is shown for each advisory. This site is neither affiliated with nor endorsed by ANSSI.
Sources for this page
- CISA KEV catalog: exploitation, date added, required action, deadline, ransomware.
- NVD (NIST): CVSS severity, publication date, description, vendor references.
- EPSS (FIRST): probability of exploitation.
- CERT-FR: advisories and alerts (in French).
- Brand and category: indicative classification by this site (method).