Fireworks Breach: Evaluation Secrets, Development Machines, and Customer Data
Fireworks AI, a hosted platform for running, evaluating and training models, disclosed unauthorized access to customer evaluation-job secrets and later development-machine activity affecting customer data. Customers can revoke exposed credentials at their issuers; the platform operator owns isolation between development systems and customer assets.
Why it relates: The breach used platform-internal credentials and development-machine authority to affect customer AI assets, a concrete provider-privilege confidentiality boundary. The attacker is an external intruder in the public account, not an established malicious employee or contractor; exact privilege grants remain unpublished.
Mappings cover unauthorized AI-platform credential and data access, affected infrastructure/data operations, supplier dependency and model-asset confidentiality. Practical overlaps explicitly preserve missing agent, MCP, model-query and completed-copy evidence. No mapping represents a verified defensive success.