Fireworks AI discloses breach after third-party cloud provider compromise
Fireworks says attackers reached its development machines through a compromised third-party cloud provider, exposing billing data and, for some customers, incomplete model weights.
According to BleepingComputer, Fireworks has disclosed a new security breach in which attackers accessed its development machines after breaching a third-party cloud provider. Fireworks said the provider hosted its research and development (R&D) machines and experienced a security incident on October 23. On October 24, the unauthorized third party used that breach to reach Fireworks development machines and run an unauthorized workload that led to additional customer data impact. Fireworks said it detected the activity the same day and closed known critical vulnerabilities.
What was exposed
The affected internal application exposed resource metadata such as Fireworks resource identifiers, display names, descriptions and creation time, along with billing information including billing and shipping addresses, invoice and payment history, account balance and credit usage. Fireworks said payment card numbers are not stored and were not impacted. For certain customers, a subset of incomplete model weights and datasets was accessed.
Fireworks said it has not detected any impact to Inference, Training SDK and Nexus, which are zero data retention (ZDR) by default. As of October 25, it had not detected further evidence of successful unauthorized access on the platform customers use.
Why it matters
Fireworks said the evidence indicates links to a financially motivated cybercriminal group that sells tokens from unauthorized access to large language models (LLMs) to illicit routing gateways, a practice known as LLMJacking. Google Threat Intelligence Group and Mandiant published a blog post about it on November 2.
According to BleepingComputer, Fireworks also notified law enforcement and customers of another incident in early October, after attackers accessed environment variables holding application programming interface (API) keys and tokens stored as secrets for a limited number of accounts. BleepingComputer reported that Google’s Threat Intelligence Group and Mandiant said LLMJacking attacks are driving record costs for compromised organizations, with cloud AI charges reaching as high as $1 million from fraudulent calls using stolen API keys.
What to do
Fireworks revoked and replaced credentials and service accounts, removed unnecessary elevated privileges, and restricted unauthorized network pathways while isolating affected systems. It also tightened access to data storage and analytics systems and disabled self-serve dataset uploads and downloads used by Batch Inference and dataset-based training while it investigates. The company said it notified law enforcement and engaged Palo Alto Networks Unit 42 for forensic analysis.
Sources
Related reading
3 postsDiscussion
Spotted an error? Send corrections to [email protected] .