CVE-2026-41267
8.1FlowiseAI · Flowise
An improper mass assignment vulnerability in the Flowise account registration endpoint allows unauthenticated attackers to modify sensitive object attributes and bypass authorization controls.
Executive summary
An unauthenticated mass assignment vulnerability in Flowise versions prior to 3.1.0 allows attackers to manipulate account metadata and escalate privileges, posing a critical risk to multi-tenant environments.
Vulnerability
This vulnerability involves improper mass assignment and authorization bypass via the account registration endpoint. An unauthenticated attacker can inject server-managed fields and nested objects, allowing them to manipulate ownership, organization associations, and role mappings.
Business impact
The ability for an unauthenticated attacker to manipulate account metadata and roles creates a significant risk of unauthorized access to sensitive large language model flows and data. In a multi-tenant environment, this flaw allows for cross-tenant data access and administrative takeovers. With a CVSS score of 8.1, this high-severity vulnerability could lead to total compromise of an organization's Flowise deployment and the underlying data processed by the models.
Remediation
Immediate Action: Update Flowise to version 3.1.0 or later immediately to resolve the mass assignment flaw.
Proactive Monitoring: Review account registration logs and user metadata for any accounts created with unexpected roles, administrative permissions, or incorrect organization associations.
Compensating Controls: Implement strict Web Application Firewall (WAF) rules to inspect incoming registration traffic for unexpected JSON fields or nested objects that deviate from the expected schema.
Exploitation status
Public Exploit Available: No (exploit_available: unknown).
Analyst recommendation
Given the high CVSS score and the nature of the vulnerability, organizations utilizing Flowise must prioritize the upgrade to version 3.1.0. This update is essential to prevent unauthorized modification of system objects and to restore the integrity of the multi-tenant authorization model. Failure to patch may result in unauthorized access to proprietary AI workflows and sensitive user data.