Fireworks AI told customers on November 9 that an intruder who entered through a third-party cloud provider accessed account and billing information, and for a limited set of accounts, secrets, datasets and incomplete model weights. The disclosure is the AI infrastructure company's second security bulletin in roughly a month, following one it published on October 9.
According to the updated bulletin, a third-party cloud provider that hosted Fireworks research and development machines had its own security incident on October 23. The next day, the unauthorized party used that access to reach Fireworks development machines and ran an unauthorized workload that led to additional customer data impact.
Fireworks said it detected the October 24 activity the same day, took remediation measures and closed known critical vulnerabilities. As of October 25, the company said it had found no further evidence of successful unauthorized access to the Fireworks platform that customers use.
Key facts
- 01Fireworks AI told customers on November 9 that an October 24 intrusion via a cloud provider exposed additional customer data.
- 02Account and billing data were affected; a limited set of accounts also had secrets, datasets and incomplete model weights accessed.
- 03Fireworks says payment card numbers are not stored on its systems and were not impacted.
- 04Inference, Training SDK and Nexus, which are zero data retention by default, show no detected impact.
- 05Fireworks links the activity to a financially motivated LLMJacking group and is working with Palo Alto Networks Unit 42.
The scope has widened since the first bulletin. At that point, Fireworks said, only secrets stored by Fireworks, such as those used in evaluation jobs, were known to have been accessed. The investigation since then turned up impacts to account and billing information.
The data in the affected internal application included resource metadata such as identifiers, display names, descriptions and creation times, along with billing and shipping addresses, invoice and payment history, account balances and credit usage. Fireworks said it does not store payment card numbers and that they were not impacted.
For certain customers, a subset of incomplete model weights and datasets was also accessed. Fireworks AI said it is notifying customers whose data was affected with details and recommended response actions, though the bulletin does not say how many accounts are involved.
The company said it has detected no impact on Inference, the Training SDK or Nexus, all of which are zero data retention by default. The investigation is ongoing, and Fireworks said it is monitoring around the clock.
Fireworks said the evidence links the activity to a financially motivated cybercriminal group that tries to sell tokens from unauthorized access to large language models to illicit routing gateways, a scheme known as LLMJacking. Google Threat Intelligence Group and Mandiant published research on LLMJacking on November 2.
The company has notified law enforcement and brought in Palo Alto Networks Unit 42 for forensics. With Unit 42, Fireworks said it revoked and replaced credentials and service accounts, removed unnecessary elevated privileges, isolated affected systems, tightened access to data storage and analytics systems, and enhanced logging and monitoring.
Fireworks has also disabled self-serve dataset uploads and downloads used by Batch Inference and dataset-based training while the investigation continues. The Training SDK remains available with zero data retention, and Fireworks Lab-based training runs in an isolated environment with added network and storage protections.
Fireworks AI, led by CEO Lin Qiao, sells companies a place to build specialized models around their own proprietary data. The company says customers whose data was affected are receiving direct notices with recommended next steps.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
