{"id":"CVE-2026-54745","aliases":["EUVD-2026-67790","GHSA-gqww-5pj5-8fq7"],"status":"published","title":"Kubeflow Pipelines: Unauthenticated SSRF and HTTP smuggling in Kubeflow Pipelines frontend /_proxy/ route, bypasses ENABLE_AUTHZ=true","description":"Kubeflow Pipelines enables users to build and deploy portable, scalable machine learning workflows. Prior to 2.17.0, the Kubeflow Pipelines frontend exposes an unauthenticated server-side request forgery vulnerability through the /_proxy/ route in frontend/server/proxy-middleware.ts. The _routePathWithReferer() function accepts an arbitrary attacker-controlled HTTP or HTTPS target and passes its origin to createProxyMiddleware without a host allowlist or filtering for loopback, link-local, RFC1918, or cluster-local addresses. The route remains outside the authorization middleware when ENABLE_AUTHZ=true and is reachable through /apis/v1beta1/_proxy/, /apis/v2beta1/_proxy/, /pipeline/apis/v1beta1/_proxy/, and /pipeline/apis/v2beta1/_proxy/, including through a crafted Referer header. Requests can forward attacker-controlled methods, headers such as Authorization, Cookie, and X-Forwarded-For, and POST bodies to reachable internal services, while returning the upstream response to the unauthenticated client. This can expose cloud metadata credentials, Kubernetes or service APIs, and other cluster-internal endpoints to unauthorized read or modification. This issue is fixed in version 2.17.0.","severity":{"version":"3.1","score":10,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H","level":"critical","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-11","score":0.00621,"percentile":0.50223,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":false},"cwes":["CWE-284","CWE-918"],"published":"2026-08-28T15:59:01Z","modified":"2026-10-28T02:30:57Z","products":[{"vendor":"kubeflow","product":"pipelines"},{"vendor":"red hat","product":"pipelines"}],"references":[{"url":"https://github.com/kubeflow/pipelines/security/advisories/GHSA-gqww-5pj5-8fq7","tags":["x_refsource_CONFIRM","exploit","advisory"]},{"url":"https://github.com/kubeflow/pipelines/pull/13511","tags":["x_refsource_MISC","fix"]},{"url":"https://github.com/kubeflow/pipelines/commit/a35f97aa4c17b25572369e5022546ab4421bdbdd","tags":["x_refsource_MISC","fix"]},{"url":"https://github.com/kubeflow/pipelines/releases/tag/2.17.0","tags":["x_refsource_MISC","web"]},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/54xxx/CVE-2026-54745.json","tags":["advisory"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-54745","tags":["advisory"]}],"sources":{"cve":[{"id":"CVE-2026-54745","title":"Kubeflow Pipelines: Unauthenticated SSRF and HTTP smuggling in Kubeflow Pipelines frontend /_proxy/ route, bypasses ENABLE_AUTHZ=true","description":"Kubeflow Pipelines enables users to build and deploy portable, scalable machine learning workflows. Prior to 2.17.0, the Kubeflow Pipelines frontend exposes an unauthenticated server-side request forgery vulnerability through the /_proxy/ route in frontend/server/proxy-middleware.ts. The _routePathWithReferer() function accepts an arbitrary attacker-controlled HTTP or HTTPS target and passes its origin to createProxyMiddleware without a host allowlist or filtering for loopback, link-local, RFC1918, or cluster-local addresses. The route remains outside the authorization middleware when ENABLE_AUTHZ=true and is reachable through /apis/v1beta1/_proxy/, /apis/v2beta1/_proxy/, /pipeline/apis/v1beta1/_proxy/, and /pipeline/apis/v2beta1/_proxy/, including through a crafted Referer header. Requests can forward attacker-controlled methods, headers such as Authorization, Cookie, and X-Forwarded-For, and POST bodies to reachable internal services, while returning the upstream response to the unauthenticated client. This can expose cloud metadata credentials, Kubernetes or service APIs, and other cluster-internal endpoints to unauthorized read or modification. This issue is fixed in version 2.17.0.","status":"published","published":"2026-08-28T15:59:01.118Z","modified":"2026-08-31T19:10:06.738Z","cvss":[{"version":"3.1","score":10,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H","level":"critical","assessor":"cna"}],"cwes":["CWE-284","CWE-918"],"updated_at":"2026-11-02T13:38:09.774038Z"}],"euvd":[{"id":"EUVD-2026-67790","description":"Kubeflow Pipelines enables users to build and deploy portable, scalable machine learning workflows. Prior to 2.17.0, the Kubeflow Pipelines frontend exposes an unauthenticated server-side request forgery vulnerability through the /_proxy/ route in frontend/server/proxy-middleware.ts. The _routePathWithReferer() function accepts an arbitrary attacker-controlled HTTP or HTTPS target and passes its origin to createProxyMiddleware without a host allowlist or filtering for loopback, link-local, RFC1918, or cluster-local addresses. The route remains outside the authorization middleware when ENABLE_AUTHZ=true and is reachable through /apis/v1beta1/_proxy/, /apis/v2beta1/_proxy/, /pipeline/apis/v1beta1/_proxy/, and /pipeline/apis/v2beta1/_proxy/, including through a crafted Referer header. Requests can forward attacker-controlled methods, headers such as Authorization, Cookie, and X-Forwarded-For, and POST bodies to reachable internal services, while returning the upstream response to the unauthenticated client. This can expose cloud metadata credentials, Kubernetes or service APIs, and other cluster-internal endpoints to unauthorized read or modification. This issue is fixed in version 2.17.0.","status":"published","published":"2026-08-28T15:59:01Z","modified":"2026-08-31T19:10:06Z","cvss":[{"version":"3.1","score":10,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H","level":"critical"}],"updated_at":"2026-11-02T18:01:19.818565Z"}],"nvd":[{"id":"CVE-2026-54745","description":"Kubeflow Pipelines enables users to build and deploy portable, scalable machine learning workflows. Prior to 2.17.0, the Kubeflow Pipelines frontend exposes an unauthenticated server-side request forgery vulnerability through the /_proxy/ route in frontend/server/proxy-middleware.ts. The _routePathWithReferer() function accepts an arbitrary attacker-controlled HTTP or HTTPS target and passes its origin to createProxyMiddleware without a host allowlist or filtering for loopback, link-local, RFC1918, or cluster-local addresses. The route remains outside the authorization middleware when ENABLE_AUTHZ=true and is reachable through /apis/v1beta1/_proxy/, /apis/v2beta1/_proxy/, /pipeline/apis/v1beta1/_proxy/, and /pipeline/apis/v2beta1/_proxy/, including through a crafted Referer header. Requests can forward attacker-controlled methods, headers such as Authorization, Cookie, and X-Forwarded-For, and POST bodies to reachable internal services, while returning the upstream response to the unauthenticated client. This can expose cloud metadata credentials, Kubernetes or service APIs, and other cluster-internal endpoints to unauthorized read or modification. This issue is fixed in version 2.17.0.","status":"published","published":"2026-08-28T20:18:17.24Z","modified":"2026-09-09T21:09:13.08Z","cvss":[{"version":"3.1","score":10,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H","level":"critical","assessor":"secondary"}],"cwes":["CWE-284","CWE-918"],"updated_at":"2026-11-02T14:12:42.366888Z"}],"osv":[{"id":"CVE-2026-54745","title":"Kubeflow Pipelines: Unauthenticated SSRF and HTTP smuggling in Kubeflow Pipelines frontend /_proxy/ route, bypasses ENABLE_AUTHZ=true","description":"Kubeflow Pipelines enables users to build and deploy portable, scalable machine learning workflows. Prior to 2.17.0, the Kubeflow Pipelines frontend exposes an unauthenticated server-side request forgery vulnerability through the /_proxy/ route in frontend/server/proxy-middleware.ts. The _routePathWithReferer() function accepts an arbitrary attacker-controlled HTTP or HTTPS target and passes its origin to createProxyMiddleware without a host allowlist or filtering for loopback, link-local, RFC1918, or cluster-local addresses. The route remains outside the authorization middleware when ENABLE_AUTHZ=true and is reachable through /apis/v1beta1/_proxy/, /apis/v2beta1/_proxy/, /pipeline/apis/v1beta1/_proxy/, and /pipeline/apis/v2beta1/_proxy/, including through a crafted Referer header. Requests can forward attacker-controlled methods, headers such as Authorization, Cookie, and X-Forwarded-For, and POST bodies to reachable internal services, while returning the upstream response to the unauthenticated client. This can expose cloud metadata credentials, Kubernetes or service APIs, and other cluster-internal endpoints to unauthorized read or modification. This issue is fixed in version 2.17.0.","status":"published","published":"2026-08-28T15:59:01.118Z","modified":"2026-10-28T02:30:57.155672429Z","cvss":[{"version":"3.1","score":10,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H","level":"critical"}],"cwes":["CWE-284","CWE-918"],"updated_at":"2026-11-02T15:45:10.818486Z"}]}}