Get ahead
VMware offers training and certification to turbo-charge your progress.
Learn moreSpring WS added destination checks to each WebServiceMessageSender, see CVE-2026-40999.
Screening a peer-supplied destination for reachability this way cannot be made correct.
A remote attacker who controls a wsa:ReplyTo or wsa:FaultTo address can still craft a destination that passes these checks yet resolves to an internal-only host, a cloud metadata endpoint, or another destination that is not otherwise reachable from the attacker.
As with the original issue, the delivery is one-way: Spring WS never reads back or exposes the destination's response, so this does not let an attacker read back that destination's contents.
This issue only affects deployments where all the following are true:
AbstractAddressingEndpointMapping subclass is registered with one or more WebServiceMessageSender instances configured for out-of-band replies, relying on the sender's default checks rather than an application-supplied DestinationPolicy.wsa:ReplyTo / wsa:FaultTo) from untrusted callers.Note that applications are affected. Configuring messageSenders on an addressing endpoint mapping is no longer enough to enable out-of-band replies.
Every WebServiceMessageSender that you configure on AbstractAddressingEndpointMapping should register a DestinationPolicy stating which destinations it may target.
Spring Web Services:
| Fix version | Availability |
|---|---|
| 5.0.3.1 | Enterprise Support Only |
| 5.0.4 | OSS |
| 4.1.5 | Enterprise Support Only |
| 4.0.20 | Enterprise Support Only |
| 3.1.10 | Enterprise Support Only |
No further mitigation steps are necessary.
The issue was identified and responsibly reported by FEARIS.
To report a security vulnerability for a project within the Spring portfolio, see the Security Policy