UpTrain, an open-source platform for evaluating generative AI applications, has an authentication flaw in version 0.7.1 and prior. The backend creates a default user with a static username that is also used as the default API key. Combined with an open Cross-Origin Resource Sharing (CORS) policy, this allows any external website to send authenticated requests by including the default API key in the 'uptrain-access-token' header. This improper authentication (CWE-287) can lead to unauthorized privileged operations on the UpTrain instance, effectively impersonating the default logged-in user. As of the publication date, no official fix or patch is available.

An attacker controlling a website can exploit the default static API key and open CORS policy to perform privileged actions on the UpTrain backend without authorization. This compromises the integrity and security of the UpTrain instance by allowing unauthorized operations as the default user. The vulnerability has a high severity rating (CVSS 7.7) due to its network accessibility, lack of authentication barriers, and potential for significant impact on the affected system.

Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until a fix is available, users should consider restricting access to the UpTrain backend, disabling or limiting CORS policies, and changing or removing the default static API key if possible. Monitoring for unusual activity and isolating the UpTrain instance from untrusted networks may reduce risk.