Fireworks AI Breach Expands: Customer Billing Data and Partial Model Weights Exposed
Fireworks AI has widened its account of a security incident first reported early last month, saying its investigation turned up additional malicious activity affecting account and billing...

Fireworks AI has widened its account of a security incident first reported early last month, saying its investigation turned up additional malicious activity affecting account and billing information, along with secrets, datasets and incomplete models for a limited set of accounts.
In a notice published October 10, 2026, the company said it had detected unauthorized use of internal Fireworks credentials that let a third party reach environment variables used in evaluation jobs in a small number of customer accounts. The activity was spotted on October 6, 2026, at 9 a.m. PDT. At the time, the AI infrastructure provider said it found only evidence that environment variables, including API keys and tokens stored as secrets, were captured in certain affected accounts, with no signs that customer models, datasets, training data, inference traffic or personal information were accessed. Fireworks revoked the compromised internal credentials and rotated service account tokens, affected Fireworks API keys and keys created by the threat actor.
The newer bulletin shows the breach may have been broader than first thought. Fireworks said attackers reached its development environment through an unnamed third-party cloud provider hosting its research and development machines on October 23, 2026. "On October 24, the unauthorized third party used it to reach Fireworks development machines and then run an unauthorized workload that led to additional customer data impact," the company said, adding that as of October 25 it had not detected further evidence of successful unauthorized access on the platform customers use.
The breach is said to have exposed resource metadata such as Fireworks resource identifiers, display names, descriptions and creation time; billing information including billing and shipping addresses, invoice and payment history, account balance and credit usage; and a subset of incomplete model weights and datasets for certain customers. Fireworks said payment card numbers are not stored by the company and were not affected, and that it is alerting customers whose data was impacted.
The company said it has carried out security hardening, including credential revocation and rotation, isolating affected systems to block further unauthorized activity and lateral movement, tightening access controls and improving logging and monitoring. "The evidence indicates links to a cybercriminal threat actor group, which is financially motivated and aimed at selling tokens from unauthorized access to large language models (LLMs) to illicit routing gateways," it said. The activity is assessed as part of an emerging threat known as LLMjacking, in which stolen credentials are used to hijack a victim's LLM and peddle access to other threat actors while the legitimate owner pays the bill. Last week, Google warned that threat actors are increasingly targeting AI infrastructure for LLMjacking across North America, South America and Asia Pacific in financial services, retail, technology and transportation.
Notion disclosed that the Fireworks AI incident affected fewer than 1% of Notion workspaces. One Reddit user wrote: "looks like all of us are part of that <1%." While the initial assessment was that the breach allowed access to Notion AI synthetic testing data, it has since emerged that potential customer data may also be involved. "Upon analysis of the new data, we discovered that it included Notion AI chat transcripts used for the Insights feature, which for the vast majority of impacted workspaces were unattributable," Notion said, explaining that the chats were not labeled with who wrote them or which workspace they came from and did not contain tool call results. Notion said no other features were affected, has disabled the Insights feature and no longer serves open-weight models through Fireworks. Affected customers are advised to review the impacted data securely inside their workspace and to rotate any sensitive information, passwords, API keys or other credentials pasted during the attack window.
Source: The Hackers News
Ready to Modernize Your Business?
Get your AI automation roadmap in minutes, not months.
Analyze Your Workflows →