AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Before you orderOffer from Amazon

Get privacy and security gear delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

Fireworks AI’s November 9 update says attackers accessed account and billing information and, for a limited set of accounts, secrets, datasets and incomplete models. Customers should review their Fireworks-connected credentials, data and billing records; the company has not said how many customers were affected or identified the third-party cloud provider.

Fireworks AI said on November 9 that attackers accessed account and billing information and, for a limited set of accounts, secrets, datasets and incomplete models, expanding the scope of a breach it first disclosed in October. The update matters to businesses that use the platform to run or fine-tune AI models because some affected material may include credentials or company data.

Fireworks’ account says the incident developed through two routes. On October 6, the company identified activity in which someone used what Fireworks called internal credentials to read environment variables from customer evaluation jobs. Such variables can contain API keys or other tokens. Fireworks said it began containment that day and started notifying affected customers on October 7.

A second event followed a security incident at an unnamed third-party cloud provider that hosted Fireworks research and development machines. Fireworks says an unauthorized party used that access on October 24 to reach development machines and run an unauthorized workload, causing additional customer data impact. The company says it detected the activity the same day and has not detected further successful access by the threat actor on the customer platform since October 25.

The company lists resource metadata and billing information among the accessible data. Billing details may include addresses, invoice and payment history, account balances and credit usage. For certain customers, Fireworks says attackers accessed a subset of datasets and incomplete model weights. It has not explained what “incomplete” means or specified which resources were involved. Fireworks says it does not store payment card numbers and says those numbers were not affected.

At a glance
updateWhen: Update published November 9, 2026; inci…
The developmentFireworks AI expanded its October breach disclosure on November 9, saying attackers reached additional customer information through a later incident involving a third-party cloud provider.
Timeline of the Fireworks AI incident: October 6, misuse of internal credentials detected, environment variables in customers' evaluation jobs read, containment starts; October 7, customers notified; October 9, employee API keys rotated and first bulletin published; October 23, a third-party cloud provider hosting Fireworks R&D machines has a security incident; October 24, the attacker reaches Fireworks development machines and runs an unauthorized workload, detected the same day; November 2 to 8, Fireworks informs Notion and confirms a dataset with Notion AI chat transcripts; November 9, highlighted, the update widens the scope to billing data, secrets, datasets and incomplete model weights for some accounts
How the incident unfolded, as Fireworks and Notion describe it. Sources: Fireworks AI security bulletins, Notion security advisory. Graphic: Thorsten Meyer AI.

Risks for AI Platform Customers

The expanded disclosure raises questions beyond the security of Fireworks’ own systems. Businesses that send data to an AI infrastructure provider may have to review what information and credentials were available to the affected systems. Exposed secrets, if present and still valid, could give access to connected services; customer teams should check their use and status rather than assume the incident was limited to billing records.

Datasets and model weights can also carry commercial value or sensitive information, depending on how a customer built and used them. Fireworks has not said what the affected datasets contained, whether the accessed material was copied or used, or whether incomplete weights could be put to use. Those unknowns limit what customers can conclude about the practical impact from the disclosure alone.

Notion has separately said that chat transcripts from its Notion AI Insights feature were in a dataset attackers may have accessed. Notion says fewer than 1% of workspaces were affected. That is a company-reported figure specific to Notion; it does not establish the number or proportion of Fireworks customers affected overall.

Products Worth Considering

Logitech MX Mechanical Wireless Illuminated Full-Size Keyboard, Clicky

Logitech MX Mechanical Wireless Illuminated Full-Size Keyboard, Clicky

Apple MacBook Pro 16.2-inch with M5 Pro, 18-core CPU, 20-core GPU, 48GB Memory, 2TB SSD, Space Black

Apple MacBook Pro 16.2-inch with M5 Pro, 18-core CPU, 20-core GPU, 48GB Memory, 2TB SSD, Space Black

As an affiliate, we earn on qualifying purchases.

From October Disclosure to Update

Fireworks’ October 9 bulletin described misuse of internal credentials and access to environment variables in a limited number of customer accounts. At that time, the company said it had not detected unauthorized access to customer models, datasets, training data, inference traffic, or personal information. Its November 9 update added impacts to account and billing information and, for some accounts, secrets, datasets and incomplete models.

The chronology in the supplied company and customer statements places the first detection on October 6, followed by customer notifications beginning October 7 and the first public bulletin on October 9. Fireworks says it rotated employee API keys on October 9 and revoked more user API keys and rotated service-account tokens on October 10. The later cloud-provider incident occurred October 23, with the unauthorized workload detected October 24.

Notion says Fireworks contacted it on November 2 about a dataset initially believed to contain synthetic test data. Fireworks told Notion on November 6 that malicious actors may have accessed potential customer data and confirmed the dataset on November 8, according to Notion’s account. Fireworks says it notified law enforcement and engaged Palo Alto Networks’ Unit 42 for forensic work.

“For a limited set of accounts, we also discovered impact to secrets, datasets, and incomplete models.”

— Fireworks AI, November 9 security update

Scope and Data Use Remain Unknown

Fireworks has not disclosed how many customers or accounts were affected, named the third-party cloud provider, or explained how the internal credentials used in the first event were obtained. The company’s update also does not identify the customers whose secrets, datasets or incomplete model weights were accessed.

The available statements do not establish whether attackers copied, retained, published or otherwise used the accessed information. Fireworks has not defined “incomplete models” or described whether the affected weights were usable. Its investigation is ongoing, and the company says it may reveal additional information.

Fireworks said it had not detected an impact to its Inference, Training SDK and Nexus products, which it describes as zero data retention by default. That statement concerns the company’s findings and product configuration; it does not identify what data was present in the affected application or settle the questions about the other exposed categories.

Customer Reviews and Investigation Updates

Customers should use Fireworks’ direct notifications and subsequent security updates to determine whether their accounts or resources were affected. As a practical response, organizations can review credentials and tokens connected to Fireworks, rotate any that may have been exposed, and check relevant access logs. They can also identify datasets, model artifacts and billing records associated with the platform, then follow their own incident procedures for sensitive information.

Notion users can consult Notion’s advisory for the company’s account of the affected workspaces and chat transcripts. Fireworks says its investigation with Unit 42 is continuing and that it has notified law enforcement. Further updates may clarify the provider involved, the number of affected customers and what happened to accessed data; those details were not established in the material available for this report.

Key Questions

What information does Fireworks say attackers accessed?

Fireworks says attackers accessed account and billing information and, for a limited set of accounts, secrets, datasets and incomplete models. It also lists resource metadata as accessible to the affected internal application.

Were payment card numbers exposed?

Fireworks says it does not store payment card numbers and that they were not impacted. The company lists other billing details, including addresses, invoices, payment history, account balances and credit usage, among the information involved.

Were Notion AI transcripts involved?

Notion says chat transcripts from its AI Insights feature were in a dataset attackers may have accessed. Notion put the share of affected workspaces at fewer than 1%; that figure does not describe Fireworks customers overall.

What should Fireworks customers check?

Customers can review Fireworks-connected API keys, tokens and access logs; identify data and model artifacts associated with affected resources; and check billing records. Fireworks has not published a complete list of affected accounts in the supplied material, so customers should also check communications from the company.

What remains unknown about the breach?

The third-party cloud provider, the total number of affected customers, how the initial internal credentials were obtained and what happened to accessed data remain unclear. Fireworks says its investigation is ongoing.

Source: Thorsten Meyer AI

You May Also Like

Actively Exploited Sandbox RCE In All Chromium Versions

Security researchers warn of active, widespread sandbox remote code execution vulnerabilities across all Chromium-based browsers, raising urgent security concerns.

‘We Hacked The FBI:’ Hackers Say They Have Data On All FBI Employees

Hackers assert they have obtained data on every FBI employee, raising security concerns. Confirmed details are limited; investigation ongoing.