GCD AI RADAR
incidents/AIR-2026-0780

CVE-2026-12798: BerriAI litellm SSRF via spec_path URL loader

Attack on AI2 Disclosed vulnerabilityInfrastructure vulnerability

BerriAI litellm contains a server-side request forgery vulnerability in its MCP OpenAPI Spec Loader. An attacker can exploit unvalidated spec_path URLs to initiate arbitrary requests remotely.

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

Details

type
Attack on AI
An AI model, agent or AI-powered product was the thing exploited, abused or leaked.
categories
  • Infrastructure vulnerability
    CVEs and advisories in AI frameworks, inference servers, MCP servers, AI developer tools and hosting stacks.
owasp llm top 10
mitre atlas
no technique for these categories
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
2 · Disclosed vulnerability
CVE or vendor advisory, patched or patchable, no confirmed exploitation.
targets
Framework / tooling
products
litellm
incident date
21 Jun 2026
first reported
21 Jun 2026
origin
GitHub Advisory Database
Vulnerability record from NVD / the GitHub Advisory Database. This product uses the NVD API but is not endorsed or certified by the NVD.
classifier confidence
0.95 · published at or above 0.80; see the methodology
tags
CVE-2026-12798 · ssrf · infra_vuln
record
created 2026-10-18 07:00Z · updated 2026-10-18 10:04Z · json

Sources · 2

  1. 2026-06-21
    CVE-2026-12798: BerriAI litellm has SSRF via Unvalidated spec_path URL in MCP OpenAPI Spec Loader
    GitHub Advisory Database · github.com/advisories/GHSA-c693-x898-5g4h
  2. 2026-06-21
    CVE-2026-12798: A weakness has been identified in BerriAI litellm up to 1.82.2.
    NVD · nvd.nist.gov/vuln/detail/CVE-2026-12798

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents