BuddyPress Instant Chat EUVD-2026-96772

Exploitation requires the plugin's chat-message submission feature to be reachable and the dual flaw present through version 1.6: the plugin does not verify the message sender belongs to the target conversation, and does not escape message content on output. No authentication is required (PR:N). The one real limiting factor is user interaction (UI:R): the payload lies dormant until a legitimate conversation member actually loads/views the affected conversation, so impact depends on a member - ideally a privileged one - opening it. No non-default configuration is needed beyond having the vulnerable plugin installed and active on WordPress with BuddyPress.

This is a credible, actionable priority rather than a high-CVSS paper tiger. The input CVSS of 8.8 (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H) describes a network-reachable, low-complexity, unauthenticated flaw requiring only that a victim later view the poisoned conversation (UI:R). A publicly available exploit is confirmed by WPScan, which materially raises the practical risk of opportunistic use. However, the input signals are partially conflicting: the C:H/I:H/A:H impact rating is aggressive for a CWE-79 stored cross-site scripting issue - standard CVSS scoring guidance for XSS yields Low confidentiality/integrity impact and a scope change (S:C) because the malicious script executes in the victim's browser (a different security authority), not full high impact on the host. I therefore assess the realistic base severity lower (CVSS 6.1) than the supplied 8.8; see assessed_cvss_vector. No EPSS score, no SSVC decision, and no CISA KEV listing were provided, so there is no evidence of active exploitation in the wild - only that exploit code exists (distinct signals that must not be conflated). Net: real priority for any site running this plugin because exploitation is unauthenticated and POC-backed, but the host-level impact implied by the input vector is overstated.

An unauthenticated attacker submits a chat message containing a malicious script payload into a conversation they do not belong to, exploiting the plugin's missing sender-membership check and missing output escaping. The script is stored and later executes in the browser of any legitimate member - potentially a moderator or administrator - who opens that conversation, enabling session/cookie theft or actions on their behalf. A public exploit for this is available via WPScan.