Red Hat OpenShift AI: Authorization bypass enables tenant credential injection
Summary
A flaw in the ai-gateway-payload-processing component of Red Hat OpenShift AI (RHOAI) lets an authenticated user bypass tenant boundaries. By specifying another tenant's model identifier in the request body or header, the attacker causes the gateway to inject that tenant's provider API keys into the request. The attacker can then make unauthorized LLM calls and consume the other tenant's resources and quotas. The flaw affects multi-tenant model serving setups; Red Hat rates it Important, and the CVSS score is 9.1 (Critical).
Affected
- Red Hat OpenShift AI (RHOAI), component ai-gateway-payload-processing; Red Hat lists the default configuration as affected
- Exact affected versions: Unknown.
Fixed in
Unknown. No vendor fix or errata found at the time of analysis; treat as unpatched.
Workaround
- No vendor-published workaround found.
- Restrict who can send requests to the model gateway and remove unnecessary user access.
- Review tenant and model mappings in the gateway configuration for cross-project model identifiers.
- If misuse is suspected, rotate the provider API keys in the shared credential store.
Exploitation
- Not listed in the CISA KEV catalogue.
- No public proof of concept reported (VulDB).
- No confirmed exploitation in the wild found.
Our recommendation
- If you run Red Hat OpenShift AI or a similar multi-tenant AI gateway, treat this as high priority: confirm exposure and restrict gateway access.
- Review usage and quota logs for model calls from unexpected projects or tenants.
- Track Red Hat errata and apply the fix as soon as it is released.
- If you do not run this stack, no immediate action is needed beyond restricting access to AI gateway credentials.
Public exploits
PoC-in-GitHub, Metasploit and Exploit-DB list nothing for this CVE.
Research sources
- Red Hat CVE page CVE-2026-117574 access.redhat.com
- Red Hat Bugzilla 2494861 bugzilla.redhat.com
- Rapid7 vulnerability database entry www.rapid7.com
- VulDB entry 428205 vuldb.com
- OffSeq threat radar entry radar.offseq.com
- Upstream project opendatahub-io/ai-gateway-payload-processing github.com
Researched and written up in English and German by an AI (Claude by Anthropic) based on NVD, CISA, EPSS and the sources listed. The vendor information is authoritative. Researched on 31 Oct 2026, 05:15.
Original description (NVD)
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. Because the service does not verify authorization before retrieving credentials from the shared store, it injects the target tenant's application programming interface (API) keys into the request. This flaw allows an attacker to perform unauthorized requests and consume another tenant's large language model (LLM) resources and quotas.
Sources
The tracker combines official, freely usable data sources – updated continuously:
- NVD (NIST) – CVEs, rating (CVSS), affected products
- CISA KEV – vulnerabilities confirmed by CISA as actively exploited
- CISA Vulnrichment – CISA assessment (SSVC) in the official CVE record
- FIRST EPSS – probability of exploitation
- Microsoft MSRC – Patch Tuesday: vulnerabilities, severity, updates
- PoC-in-GitHub – public proof-of-concept repositories
- Metasploit (Rapid7) – modules of the Metasploit Framework
- Exploit-DB (OffSec) – archive of public exploits (links only)
- Patch days: official vendor pages (linked per vendor) (Vendor patch days)
This product uses data from the NVD API but is not endorsed or certified by the NVD. EPSS data © FIRST.org.
Affected? We can help.
We check whether your systems are affected, install updates and close the gap – continuously with the System Health Checker if you wish.