Fireworks AI Says Breach Exposed Customer Billing Data, Secrets and Incomplete Models
Summary
Fireworks AI says attackers accessed its development environment and affected a limited set of customer accounts, exposing account and billing information, secrets, datasets and incomplete models. The company revoked credentials and hardened systems.
Key points
- Fireworks AI said unauthorized use of internal credentials exposed environment variables, including API keys and tokens stored as secrets, in some customer accounts.
- The company later reported that attackers accessed its development environment through an unnamed third-party cloud provider and ran an unauthorized workload.
- Fireworks AI said the broader incident affected account and billing information, secrets, datasets and incomplete models for a limited set of accounts.
- The company said payment card numbers were not stored by Fireworks AI and were not affected, and that it was notifying impacted customers.
- Fireworks AI reported revoking and rotating credentials and API keys, isolating affected systems, tightening access controls, and improving logging and monitoring.
- The company said evidence linked the activity to a financially motivated cybercriminal group seeking to sell unauthorized LLM access; it assessed the activity as LLMjacking.
Article Details
- Victim Organization
- Fireworks AI
- Incident Type
- Unauthorized access using internal Fireworks credentials, followed by access to development machines and execution of an unauthorized workload that affected customer data.
- Incident Date
- 2026-10-06
- Disclosure Date
- 2026-10-10
- Data Types Exposed
- Account information
- Billing information
- Secrets, including API keys and tokens stored as environment variables
- Datasets
- Incomplete models
- Operational Impact
- The activity detected on October 6 involved unauthorized access to environment variables in a limited number of customer accounts. Further activity reached Fireworks development machines on October 23, and an unauthorized workload ran on October 24, leading to additional customer data impact. Fireworks said that as of October 25 it had detected no further evidence of successful unauthorized access on its customer platform. It revoked and rotated credentials, isolated affected systems, tightened access controls, and improved logging and monitoring.
- Ransom Or Extortion
- No ransom demand was reported. Fireworks said evidence indicated links to a financially motivated cybercriminal group aiming to sell tokens obtained through unauthorized access to LLMs to illicit routing gateways.
- Claim Status
- confirmed