Executive brief
IBM Langflow OSS, a tool used for building AI applications, contains a critical security flaw in how it protects sensitive information. The system uses a weak encryption method to store credentials like API keys, database passwords, and login tokens. If an attacker obtains the application's secret key file, they can easily decrypt and steal all stored credentials, potentially leading to full unauthorized access to connected third-party services and databases.
Technical details
IBM Langflow OSS (versions 1.0.0 to 1.10.0) utilizes a cryptographically weak Pseudo-Random Number Generator (PRNG), specifically Python's Mersenne Twister, to derive Fernet encryption keys for credentials stored in the database. The derivation process seeds the PRNG with the application's SECRET_KEY; if this key is shorter than 32 characters, the resulting encryption key is fully deterministic and reproducible. Furthermore, for keys 32 characters or longer, the raw SECRET_KEY was used directly as the Fernet key. When combined with other vulnerabilities like path traversal to exfiltrate the secret_key file, an attacker can decrypt all stored API keys, database passwords, and OAuth tokens offline. The issue is resolved in version 1.10.1.
Affected products
- IBM Langflow OSS 1.0.0 through 1.10.0
Timeline
- 2026-06-29: advisory: Initial publication of IBM security bulletin
- 2026-06-30: disclosed: NVD publication date