CVE-2026-41276

CVE-2026-41276 is an authentication bypass vulnerability affecting FlowiseAI Flowise, an open-source drag-and-drop user interface for building customized large language model flows, in versions prior to 3.1.0. The flaw resides in the resetPassword method of the AccountService class, where no validation occurs to confirm that a password reset token has been generated for the target user account. By default, the reset token value stored in user accounts is null or an empty string, particularly if the user has previously reset their password.

Remote attackers can exploit this vulnerability without authentication by knowing only the target user's email address. They submit a request to the /api/v1/account/reset-password endpoint with a null or empty string as the reset token value, allowing them to overwrite the user's password with one of their choosing. This enables full account takeover, granting attackers high confidentiality, integrity, and availability impacts as reflected in the CVSS v3.1 base score of 9.8 (AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H), linked to CWE-287 (Improper Authentication).

The GitHub security advisory (GHSA-f6hc-c5jr-878p) confirms the issue is resolved in Flowise version 3.1.0, recommending that users upgrade immediately to mitigate the risk. No additional workarounds are specified in the provided references.

Flowise's role in constructing LLM-based workflows introduces AI/ML relevance, as compromised instances could potentially allow attackers to manipulate or exfiltrate custom model flows. No evidence of real-world exploitation is noted in the available details.