Executive brief
Flowise is an open-source chatbot framework that enables users to build and deploy conversational AI applications. The text-to-speech feature allows chatbots to generate audio responses. A vulnerability in the abort endpoint allows unauthenticated attackers to terminate active chatbot conversations for any user by sending specially crafted requests, causing targeted service disruption and incomplete responses to end users.
Technical details
The /api/v1/text-to-speech/abort endpoint is whitelisted (requires no authentication) and accepts user-supplied chatflowId and chatId parameters from the request body without verifying ownership or authorization. The handler aborts both the text-to-speech controller and the main chatflow AbortController for a prediction session, constructed as "${chatflowId}_${chatId}". An unauthenticated attacker who knows a chatflowId (obtainable from widget source or URLs) and a chatId (often visible in client-side JavaScript or network traffic) can call this endpoint to terminate any in-progress chatflow prediction. There is no rate limiting, enabling continuous abuse to persistently disrupt a user's session. The vulnerability is fixed in version 3.1.4 by adding authorization checks to verify ownership of the chatflow and chat session.
Affected products
- FlowiseAI Flowise before 3.1.4
Timeline
- 2026-08-28: disclosed
- 2026-09-12: patched: Version 3.1.4 released