Security / News explained
A FreeSWITCH message should not become an admin command
CISA’s November 9 listing renews attention to a FreeSWITCH SIP MESSAGE flaw. The advisory explains why ordinary message authentication and a tighter inbound ACL do not close the command-execution route.
CISA added FreeSWITCH vulnerability CVE-2026-116779 to its Known Exploited Vulnerabilities catalog on November 9, 2026. The new listing gives operators a current reason to revisit a previously disclosed flaw in the boundary between message delivery and administrative commands. [2]
The October 28 advisory lists versions through 1.11.2 as affected and 1.11.3 or later as patched, and reports observed exploitation. [1]
A message reached the wrong kind of interface
An accepted inbound SIP MESSAGE could select a chat interface that dispatched administrative commands. The advisory describes execution under service privileges, secret disclosure and call control. [1]
Zyel’s interpretation is that this is an authorization-boundary problem hiding inside a messaging feature. The service was accepting a message for delivery while also allowing message input to choose a more powerful destination. A system needs a deliberate decision when an ordinary input crosses into administrative authority.
Consider a hypothetical communications platform where a handset account is allowed to exchange messages. That account’s legitimate role does not explain why it should be allowed to alter server behavior. A review that asks only whether the sender has an account can miss the separate question of what that account is authorized to make the server do.
The advisory covers profiles accepting inbound MESSAGE, enabled in stock internal and external profiles. It says the inbound ACL does not cover MESSAGE. Authentication may require an ordinary account; stock empty-password directory users can remove the need for an attacker’s own account under the described conditions. [1]
A control has to cover the relevant route
The distinction matters for a change meeting. A statement such as “we tightened the access list” describes an action. To assess the action, the reviewer needs to know whether that control actually governs the message type involved. A security setting can be correctly applied and still answer a different question from the one the vulnerability raises.
In a hypothetical response, one engineer reviews network exposure while another reviews extension accounts. Both tasks may be useful. The response owner still needs a decision about the route from an accepted message to an administrative operation. Making that route explicit gives the two reviews a shared target and reduces the chance of declaring completion from unrelated evidence.
The project says MESSAGE authentication alone is insufficient. Disabling the system-command API alone also leaves other APIs reachable. Its workaround is disabling inbound SIP chat on profiles that do not need it. [1]
Our practical reading is that a temporary restriction needs an owner and an understood service effect. Removing an unused feature can be straightforward; removing a feature someone depends on requires an operational decision. The advisory supplies the technical boundary, while the organization has to identify the business function that boundary currently supports.
The fixed release adds an explicit gate
The 1.11.3 release notes list a new enable-chat-api-proto setting in the SIP and Verto modules, alongside a bound on parsing the inbound MESSAGE recipient. The release is presented as containing critical security fixes and stability improvements. [3]
The advisory says the per-profile gate defaults off, blocking MESSAGE selection of the administrative interface while preserving other chat routes. Enabling it reopens that capability. [1]
For a team planning the update, our recommendation is to make the intended permission boundary part of acceptance. The question is whether ordinary communication still works while administrative behavior stays limited to its approved route. An approved, bounded verification can address that question without turning the production change into an exploit demonstration.
A useful completion record could identify the running release, the relevant profiles, the intended gate state and the ordinary communication checks performed. Exposure assessment is a separate decision: which routes were reachable, what evidence was reviewed and whether any uncertainty remains. These are suggested review criteria, not findings from a customer system.
The lasting engineering lesson is to keep delivery permission and administrative permission distinct wherever input selects an internal operation. A messaging account should have a clearly bounded job. When a service can explain that boundary, the operator can assess both the repair and the legitimate behavior that must survive it.
Sources + further reading
Source dates describe the original material. Publication above is the date this Zyel article went live.
- FreeSWITCH advisory: inbound SIP MESSAGE administrative command executionSignalWire / FreeSWITCH project · Checked Nov 10, 2026
- CISA Known Exploited Vulnerabilities catalogCISA · Checked Nov 10, 2026
- FreeSWITCH v1.11.3 release notesSignalWire / FreeSWITCH project · Checked Nov 10, 2026
Reader discussion
What would you ask next?
Questions, useful corrections and thoughtful disagreement are welcome. Comments are reviewed before they appear.
No reader comments yet.