A post on a dark web forum alleges that an unsecured TensorFlow database may be exposing sensitive files, including source code. While TensorFlow, an open-source machine learning framework developed by Google, publicly shares its codebase, the threat actor claims that the exposed files include content not found in public repositories.
According to the forum post, the database relies on Google’s API for file uploads but lacks password protection. The threat actor asserts that anyone with access to the link can download the stored files. However, the actual value and sensitivity of the leaked content remain uncertain.
Despite TensorFlow’s open-source nature, the post further alleges that some of the files contain passwords, raising concerns over potential security risks.
