Together AI discloses prompt sample leak incident
Understand this story
Together AI has disclosed a security incident: in June and July 2026, some prompt samples in its serverless inference service were downloaded by unauthorized parties, and the company is notifying customers about this. Together AI said it has made zero data retention (ZDR) the default setting for new accounts; for customers handling sensitive prompt data, the company recommends enabling ZDR in their account settings.
AI overview based on source reports · Updated 36 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Together AI BlogSecurity event: Prompt leakage for requests in June and July of 2026
Together AI disclosed that an unauthorized party downloaded stored prompt samples from its serverless inference service after an AI coding agent committed a credential to the public GitHub repository xorl, exposing a private Hugging Face 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, and the party searched them for credentials to other services, testing five canary credentials tied to a Dragos security experiment. Together AI says it has strengthened credential and development controls, made zero data retention the default for new accounts, and is continuing to review the affected data.
Activity for this story
Not enough continuous observations to show a trend yet.