CVE-2026-59821: LiteLLM Custom Code Guardrails Bypass Sandboxing
LiteLLM
LiteLLM custom code guardrail creation paths failed to apply proper sandboxing and validation mechanisms. This flaw allowed privileged users to execute arbitrary Python code in the proxy container.
disclosed CVENo confirmed exploitation, so this record is tracked as a disclosed vulnerability and not counted as an incident. It moves into the counts if CISA or the vendor confirms exploitation. Why
- CVSS
- 2.1
- EPSS
- 0.01 · top 40% · 2026-11-13
- fix status
- Fix status unknown
- first report
- 8 Jul 2026
- last update
- 22 Jul 2026
- CVE ids
Timeline · 2
Sources · 2
Advisories and databases
- 2026-07-08CVE-2026-59821: LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format.NVD · nvd.nist.gov/vuln/detail/CVE-2026-59821
- 2026-07-22CVE-2026-59821: LiteLLM: Custom Code Guardrails production endpoints bypass code safety checksGitHub Advisory Database · github.com/advisories/GHSA-72m8-9m7m-h278
Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.
Classification
- lane
- Patch now
- AI role
- Attack on AI
- categories
- Infrastructure vulnerability
- severity
- 2 · Disclosed vulnerability
- fix status
- Fix status unknown
- targets
- Framework / tooling
- owasp llm top 10
- LLM03:2025 Supply Chain
- mitre atlas
- no technique for these categories
- dates
- happened 22 Jul 2026 · first reported 8 Jul 2026
- origin
- GitHub Advisory DatabaseVulnerability record from NVD / the GitHub Advisory Database. This product uses the NVD API but is not endorsed or certified by the NVD.
- record
- confidence 0.95 · created 2026-10-18 07:00Z · updated 2026-10-18 10:04Z
Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned
Copy a brief
This record as a short Markdown brief, for a ticket, a chat or a coding agent.
the brief as text
# CVE-2026-59821: LiteLLM Custom Code Guardrails Bypass Sandboxing https://gcdradar.vercel.app/incidents/cve-2026-59821-litellm-custom-code-guardrails-bypass-sandbox · AIR-2026-0848 · closed 2026-07-08 - Lane: Patch now - Severity: 2 (Disclosed vulnerability) - Fix status: Fix status unknown - First reported: 2026-07-08 (happened 2026-07-22) - AI role: Attack on AI - Products: LiteLLM - CVE ids: CVE-2026-59821 - CVSS: 2.1 ## Summary LiteLLM custom code guardrail creation paths failed to apply proper sandboxing and validation mechanisms. This flaw allowed privileged users to execute arbitrary Python code in the proxy container. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2026-07-22: CVE-2026-59821: LiteLLM: Custom Code Guardrails production endpoints bypass code safety checks (https://github.com/advisories/GHSA-72m8-9m7m-h278) - 2026-07-08: CVE-2026-59821: LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. (https://nvd.nist.gov/vuln/detail/CVE-2026-59821) ## Sources - 2026-07-08 NVD: CVE-2026-59821: LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. https://nvd.nist.gov/vuln/detail/CVE-2026-59821 - 2026-07-22 GitHub Advisory Database: CVE-2026-59821: LiteLLM: Custom Code Guardrails production endpoints bypass code safety checks https://github.com/advisories/GHSA-72m8-9m7m-h278 Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closed9.8LiteLLM CVE-2026-30623: Remote code execution vulnerability in LiteLLM
Patch now · Disclosed · since 15 Jul 2026 · CVE-2026-30623
AI - closed2.1LiteLLM CVE-2026-59819: LiteLLM Local File Read Vulnerability
Patch now · Disclosed · since 22 Jul 2026 · updated 22 Jul 2026 · CVE-2026-59819
AI - closed6.1LiteLLM CVE-2026-59820: LiteLLM Arbitrary File Write via Skills Archive
Patch now · Disclosed · since 22 Jul 2026 · updated 22 Jul 2026 · CVE-2026-59820
AI - closed8.8LiteLLM CVE-2026-59822: LiteLLM MCP authentication bypass via OAuth2 fallback
Patch now · Disclosed · since 22 Jul 2026 · updated 22 Jul 2026 · CVE-2026-59822
AI - closed7.3LiteLLM CVE-2026-12795: LiteLLM SSO Debug Flow Authentication Bypass Vulnerability
Patch now · Disclosed · since 21 Jun 2026 · CVE-2026-12795
AI - closed5.4LiteLLM CVE-2026-12770: LiteLLM Admin Key Handler Improper Authorization Vulnerability
Patch now · Disclosed · since 21 Jun 2026 · CVE-2026-12770
AI