Executive brief
Discourse is an open-source platform used for hosting online discussion forums and communities. A security flaw in the AI conversation sharing feature allows an attacker to embed malicious code into conversation titles. If another user views a preview of that conversation, the code could execute in their browser, potentially allowing the attacker to hijack their session or perform actions on their behalf.
Technical details
A stored cross-site scripting (XSS) vulnerability exists in the discourse-ai plugin within the Discourse platform. The issue stems from improper neutralization of input in shared AI conversation titles when rendered in 'onebox' previews. An authenticated attacker can craft a conversation title containing malicious HTML or JavaScript payloads. When other users view the onebox preview of the shared conversation, the payload executes in their browser context. This can lead to session hijacking or unauthorized actions performed on behalf of the victim. The vulnerability has been patched in versions 2026.1.3, 2026.2.2, and 2026.3.0.
Affected products
- Discourse Discourse 2026.1.0-latest to 2026.1.2, 2026.2.0-latest to 2026.2.1, 2026.3.0-latest
Timeline
- 2026-03-31: advisory: Vendor advisory GHSA-pjc5-8x3w-rfwx published
- 2026-03-31: disclosed: CVE-2026-32243 published to NVD
- 2026-03-31: patched: Fix committed to Discourse repository