ai-gateway-payload-processing: Unauthorized resource access via model identifier (CVE-2026-117574)
An authenticated remote attacker can exploit a flaw in ai-gateway-payload-processing by submitting a request specifying another tenant's model identifier, resulting in unauthorized LLM resource consumption.
CVE search metadata
CVE search record: CVE-2026-117574. Severity: critical. CVSS: 9.1. KEV: no. Product: ai-gateway-payload-processing. Brief: ai-gateway-payload-processing: Unauthorized resource access via model identifier (CVE-2026-117574). Brief link: https://feed.craftedsignal.io/briefs/2026-10-ai-gateway-payload-processing-auth-bypass/
Triage facts
Revision 1- Applicability
- ai-gateway-payload-processing
- Evidence
- Unknown
- Severity
- Critical
- Sources
- 1 public source
- Freshness
- Revision 1Oct 28, 20:37 UTC
- Vulnerabilities
- 1 CVE
- Source quality
- A2Roughly even chance
A flaw was found in ai-gateway-payload-processing. An authenticated remote attacker can exploit this vulnerability by submitting a request specifying another tenant's model identifier in the request body or header. The service does not verify authorization before retrieving credentials from the shared store, causing it to inject the target tenant's application programming interface (API) keys into the request. This allows an attacker to perform unauthorized requests and consume another tenant's large language model (LLM) resources and quotas.