From the source
Together AI is notifying customers that an unauthorized party downloaded stored prompt samples from its serverless inference service after a credential for a private repository was exposed in a public code repository.
- The samples came from requests to the
MiniMax M3, GLM-5.2, and Kimi K2.6 models, primarily on June 24, July 4, July 8, and July 14, 2026.
- The prompts had been collected as part of routine software testing and stored in a private
Hugging Face repository.
- During development of the public
GitHub repository xorl, an AI coding agent committed a credential that provided access to that repository; automated scanning detected the exposure and the commit was reverted, but the unauthorized party obtained the credential during the exposure window.
- The party then searched the samples for credentials to other services and tested whether they were valid, and five of the credentials tested were canary credentials associated with a Dragos security experiment.
- Together AI says it has no reason to believe the data was leaked further and understands it has now been deleted, and that it has worked with an external forensic investigation firm and law enforcement.
- The company says it has since adopted scoped or short-lived credentials, checks all pull requests for credentials, made zero data retention the default for new accounts, and updated its development processes for AI coding agents.






