Spring Projects: Three High-Severity CVEs Disclosed Across Cloud Gateway, AI Vector Stores, and Sleuth
Spring Projects disclosed three high-severity CVEs on June 15, 2026, spanning Spring Cloud Gateway, Spring AI Vector Stores, and Spring Cloud Sleuth, with patches now available.
Key findings
- Three high-severity CVEs disclosed together on June 15, 2026 across Spring Cloud Gateway, Spring AI, and Spring Cloud Sleuth
- CVE-2026-47825 (CVSS 8.6) allows IP spoofing via untrusted X-Forwarded-For headers in Spring Cloud Gateway
- CVE-2026-47835 (CVSS 8.6) enables arbitrary query injection in Spring AI Vector Stores for Elasticsearch, OpenSearch, and GemFire
- CVE-2026-41708 (CVSS 7.5) causes denial-of-service through crafted calls in Spring Cloud Sleuth instrumentation
- All three vulnerabilities have patches available in the respective component releases
- No active exploitation reported at the time of disclosure
On June 15, 2026, the Spring Projects team disclosed three high-severity vulnerabilities across Spring Cloud Gateway, Spring AI Vector Stores, and Spring Cloud Sleuth. The batch, published within a one-hour window, covers a proxy-header trust bypass, a vector-store injection flaw, and a denial-of-service condition — each rated between 7.5 and 8.6 on the CVSSv3 scale. Organizations running any of these components should prioritize patching immediately.
**CVE-2026-47825** (CVSS 8.6) affects Spring Cloud Gateway Server in both WebMVC and WebFlux variants. Under certain configuration scenarios, the gateway forwards X-Forwarded-For and Forwarded headers from untrusted proxies, allowing an attacker to spoof the originating client IP. This can subvert IP-based access controls, rate-limiting, and audit logging. The fix is included in Spring Cloud Gateway 3.1.13 and 4.1.13.
**CVE-2026-47835** (CVSS 8.6) targets Spring AI Vector Stores, specifically the Elasticsearch, OpenSearch, and GemFire VectorDB store implementations. By injecting special characters into queries, an attacker can force the execution of arbitrary queries against the underlying vector database. This affects spring-ai-elasticsearch-store, spring-ai-opensearch-store, and spring-ai-gemfire-store in Spring AI versions 1.0.0 through 1.0.x. The patch is available in Spring AI 1.0.1.
**CVE-2026-41708** (CVSS 7.5) resides in Spring Cloud Sleuth instrumentation. A specially crafted call can trigger a denial-of-service condition when the application uses a vulnerable version of spring-cloud-sleuth-instrumentation and Spring TX instrumentation is not disabled. The advisory recommends upgrading to the latest patched release of Spring Cloud Sleuth.
All three CVEs were addressed in coordinated advisories published by the Spring team on the same day. Users are urged to upgrade to the fixed versions listed above. No evidence of in-the-wild exploitation has been reported at the time of disclosure.
This batch underscores the breadth of the Spring ecosystem and the importance of keeping even auxiliary components — such as gateway header handling, AI vector-store integrations, and tracing instrumentation — up to date. Organizations using Spring Cloud Gateway, Spring AI, or Spring Cloud Sleuth should review their dependency trees and apply the relevant patches.