How severe is CVE-2026-12797?

CVE-2026-12797 has a CVSS 3.x base score of 6.3, rated medium severity. It is exploitable over network with low attack complexity, requires low privileges and no user interaction. Impact on confidentiality is low, integrity low, and availability low.

Is CVE-2026-12797 being actively exploited?

It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 0% (13th percentile), an estimate of the probability of exploitation in the next 30 days.

What products are affected by CVE-2026-12797?

CVE-2026-12797 affects Litellm. See the affected-products list for the exact vulnerable versions.

How do I fix CVE-2026-12797?

Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround.

Does CVE-2026-12797 have an EU (EUVD) identifier?

Yes. CVE-2026-12797 is tracked in the ENISA EU Vulnerability Database (EUVD) as EUVD-2026-38156.

When was CVE-2026-12797 published?

CVE-2026-12797 was published on 2026-06-21 and last updated on 2026-06-24.

CVE-2026-12797

CVE-2026-12797 is a medium-severity vulnerability in Litellm with a CVSS 3.x base score of 6.3. It is not currently listed as actively exploited by CISA, and its EPSS exploit-prediction score is low. The underlying weakness is classified as CWE-285.

Key facts

Severity: Medium (CVSS 3.x base score 6.3)
CVSS v2: 6.5
CVSS v4: 2.1
EPSS exploit prediction: 0% (13th percentile)
Actively exploited: Not listed in CISA KEV
EU (EUVD) id: EUVD-2026-38156
Weakness: CWE-285
Affected product: Litellm
Published: 2026-06-21
Last modified: 2026-06-24

Description

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.

Frequently asked questions

What is CVE-2026-12797?: 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.
How severe is CVE-2026-12797?: CVE-2026-12797 has a CVSS 3.x base score of 6.3, rated medium severity. It is exploitable over network with low attack complexity, requires low privileges and no user interaction. Impact on confidentiality is low, integrity low, and availability low.
Is CVE-2026-12797 being actively exploited?: It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 0% (13th percentile), an estimate of the probability of exploitation in the next 30 days.
What products are affected by CVE-2026-12797?: CVE-2026-12797 affects Litellm. See the affected-products list for the exact vulnerable versions.
How do I fix CVE-2026-12797?: Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround.
Does CVE-2026-12797 have an EU (EUVD) identifier?: Yes. CVE-2026-12797 is tracked in the ENISA EU Vulnerability Database (EUVD) as EUVD-2026-38156.
When was CVE-2026-12797 published?: CVE-2026-12797 was published on 2026-06-21 and last updated on 2026-06-24.

References

Affected products (1)

cpe:2.3:a:litellm:litellm:*:*:*:*:*:*:*:*

More vulnerabilities in Litellm

CVE-2026-49468 — Critical (CVSS 9.8): LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, a Host-header…
CVE-2026-42208 — Critical (CVSS 9.8): LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.81.16 to before…
CVE-2024-5751 — Critical (CVSS 9.8): BerriAI/litellm version v1.35.8 contains a vulnerability where an attacker can achieve remote code execution. The…
CVE-2024-2952 — Critical (CVSS 9.8): BerriAI/litellm is vulnerable to Server-Side Template Injection (SSTI) via the `/completions` endpoint. The…
CVE-2026-35030 — Critical (CVSS 9.1): LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.0, when JWT…
CVE-2026-47102 — High (CVSS 8.8): LiteLLM prior to 1.83.10 allows a user to modify their own user_role via the /user/update endpoint. While the endpoint…

All CVEs affecting Litellm →

Other CWE-285 (Improper Authorization) vulnerabilities

CVE-2025-65041 — Critical (CVSS 10.0): Improper authorization in Microsoft Partner Center allows an unauthorized attacker to elevate privileges over a network.
CVE-2023-33189 — Critical (CVSS 10.0): Pomerium is an identity and context-aware access proxy. With specially crafted requests, incorrect authorization…
CVE-2022-2595 — Critical (CVSS 10.0): Improper Authorization in GitHub repository kromitgmbh/titra prior to 0.79.1.
CVE-2021-28799 — Critical (CVSS 10.0): An improper authorization vulnerability has been reported to affect QNAP NAS running HBS 3 (Hybrid Backup Sync. ) If…
CVE-2026-34048 — Critical (CVSS 9.9): Coolify is an open-source and self-hostable tool for managing servers, applications, and databases. Prior to…
CVE-2026-5412 — Critical (CVSS 9.9): In Juju versions prior to 2.9.57 and 3.6.21, an authorization issue exists in the Controller facade. An authenticated…

Browse all CWE-285 (Improper Authorization) vulnerabilities →