CVE-2026-42208
CVE-2026-42208 is a critical-severity vulnerability (CVSS 4.0 CVSS-B score 9.3). BerriAI LiteLLM contains a SQL injection vulnerability that allows an attacker to read data from the proxy's database and potentially modify it, leading to unauthorized access to the proxy and the credentials it manages. CISA added it to the Known Exploited Vulnerabilities catalog on 2026-05-08, with a remediation due date of 2026-05-11 for US federal civilian agencies. As of 2026-11-22, EPSS estimates an 89.4% probability of exploitation in the next 30 days, in the 99th percentile of scored CVEs.
Updated
- CVSS 4.0
- 9.3 CRITICAL
- EPSS
- 89.4% (99th pct)
- CISA KEV
- Exploited
- Published
- 2026-05-08
- Weakness
- CWE-89
BerriAI LiteLLM SQL Injection Vulnerability is confirmed as actively exploited and is listed in the CISA KEV catalog.
Required action
Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.
CISA notes
https://github.com/BerriAI/litellm/security/advisories/GHSA-r75f-5x8p-qvmc ; https://nvd.nist.gov/vuln/detail/CVE-2026-42208
Description
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.81.16 to before version 1.83.7, a database query used during proxy API key checks mixed the caller-supplied key value into the query text instead of passing it as a separate parameter. An unauthenticated attacker could send a specially crafted Authorization header to any LLM API route (for example POST /chat/completions) and reach this query through the proxy's error-handling path. An attacker could read data from the proxy's database and may be able to modify it, leading to unauthorised access to the proxy and the credentials it manages. This issue has been patched in version 1.83.7.
Key Findings
as of 2026-11-22- CISA added CVE-2026-42208 to the KEV catalog 1 day after its CVE record was published. The median for entries added in 2026 so far is 12.5 days.
- It is one of 4 BerriAI entries in the CISA KEV catalog, and one of 4 added in 2026.
- Its current EPSS score (89.4%) is higher than that of 71% of CISA KEV entries.
Severity
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:NOpen in CVSS calculator Remediation Timeline
BOD 26-04, US federal agenciesDeadlines for US federal civilian agencies under Table 1 of CISA BOD 26-04, from the CVE's KEV status (listed) and CISA's SSVC values (automatable yes, technical impact total, 2026-05-08).
How the timeline is set
Table 1 of CISA's Binding Operational Directive 26-04 sets the time US federal civilian agencies have to remediate a vulnerability. It depends on four inputs: whether the asset is publicly exposed, whether the CVE is in the KEV catalog, whether an adversary can automate its exploitation, and whether exploitation gives partial or total control of the asset, which SSVC calls technical impact. CISA publishes the last three for every CVE through services such as its Vulnrichment program.
Days are calendar days. The clock starts at whichever comes first: CISA adds the CVE to the KEV catalog, or the agency identifies the vulnerability on an asset and reports it to its Continuous Diagnostics and Mitigation (CDM) dashboard. A publicly exposed asset is accessible to unauthenticated or untrusted entities via public networks, such as the internet. Forensic triage means assessing the asset to determine whether it has been compromised.
CISA calculates the KEV due date with the same table, using its own evidence on whether the vulnerability is present in a publicly exposed asset. Each agency makes the final exposure decision for each of its assets.
Affected Products
| Vendor | Product | Affected versions |
|---|---|---|
| BerriAI | BerriAI litellm | >= 1.81.16, < 1.83.7 |
References
2 linksRecord Details
- Published
- 2026-05-08
- Last updated
- 2026-07-15
- Assigner (CNA)
- GitHub_M
- Credited to
- -
More from BerriAI
LiteLLM: Authenticated command execution via MCP stdio test endpoints
BerriAI LiteLLM Unprotected Alternate Channel Vulnerability
LiteLLM: MCP Authentication Bypass via OAuth2 Passthrough Fallback
LiteLLM affected by privilege escalation via unrestricted proxy configuration endpoint
CVE record data © The MITRE Corporation, used under the CVE Terms of Use. Exploited-in-the-wild data from the CISA Known Exploited Vulnerabilities catalog (CC0). Exploit prediction scores by FIRST EPSS (first.org/epss). Alert titles and links from CISA cybersecurity alerts. This site is not affiliated with or endorsed by MITRE, NVD, CISA, or FIRST.