🔍 Read the full analysis: What To Ask Your Team About The Fireworks AI Breach on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Fireworks AI’s November 9 update says an incident first disclosed in October also affected account and billing information and, for a limited set of accounts, secrets, datasets and incomplete model weights. Notion says chat transcripts from its AI Insights feature may have been in an accessed dataset, affecting fewer than 1% of workspaces. The number of customers affected, the identity of the cloud provider and what happened to accessed data remain unclear.
Fireworks AI said on November 9 that a security incident it first disclosed in October reached further than initially reported, affecting account and billing information and, for a limited set of accounts, secrets, datasets and incomplete models. The company described two routes into its systems, including an October 24 intrusion through a third-party cloud provider; the update matters to organizations that use Fireworks to run or fine-tune AI models.
In its October 9 bulletin, Fireworks said an attacker had used internal credentials to read environment variables from a limited number of customer evaluation jobs. Those variables can contain sensitive settings such as API keys and tokens. At the time, the company said it had not detected access to customer models, datasets, training data, inference traffic or personal information. Its November 9 update changed that account: Fireworks said it later discovered impacts to account and billing information, along with secrets, datasets and incomplete models for a limited set of accounts.
Fireworks says the later access followed an incident at a third-party cloud provider that hosted its research and development machines. According to the company, an unauthorized party used that access on October 24 to reach development machines and run an unauthorized workload that led to additional customer data impact. Fireworks says it detected the activity that day and has not detected further successful unauthorized access to the customer-facing platform since October 25. It has not named the provider.
The company says accessed information included resource metadata, billing and shipping addresses, invoice and payment history, account balances and credit usage. For certain customers, it says a subset of datasets and incomplete model weights was accessed. Fireworks says it does not store payment card numbers and they were not affected. It also says it has detected no impact to Inference, Training SDK or Nexus, which it describes as zero-data-retention by default. The company says it notified law enforcement and hired Palo Alto Networks Unit 42 to assist with forensic work.
Notion separately said chat transcripts from its Notion AI Insights feature were in a dataset that attackers may have accessed. Notion put the share of potentially affected workspaces at fewer than 1%. That statement identifies a possible exposure involving one customer’s data; it does not establish that attackers read or used every transcript in the dataset.
What to Ask Your Team About the Fireworks AI Breach
Fireworks AI’s November update expanded the reported scope of an October incident. Customers should confirm what data and systems may have been affected in their own accounts.
What may be in scope
The update names additional information that customers should check against their own records.
Billing and usage records
Reported information includes billing and shipping addresses, invoices, payment history, account balances and credit usage.
Secrets, datasets and models
For a limited set of accounts, Fireworks reported access to secrets, datasets and incomplete model weights.
AI Insights transcripts
Notion says transcripts were in a dataset attackers may have accessed; this does not establish that every transcript was read.
Questions for your team
Review your use of Fireworks and gather account-specific evidence.
Were evaluation jobs, datasets, fine-tuned models or secrets in the potentially affected environment?
Do your logs show access or changes to relevant data, jobs or billing records during the period?
Ask Fireworks whether your account was affected and what action, if any, is recommended.
What is known and unresolved
The company’s update describes its findings so far; key details remain open.
- Activity involving a third-party provider’s R&D machines was detected on October 24.
- No impact detected to Inference, Training SDK or Nexus.
- Payment card numbers are not stored and were not affected.
- Which customers and accounts were affected, and how many.
- The cloud provider’s identity and the meaning of “incomplete” model weights.
- Whether accessed data was copied, shared or used.
Disclosure timeline
Questions for Fireworks Customers
The expanded disclosure changes what customers need to review. Organizations should ask whether they had evaluation jobs, datasets, fine-tuned models or secrets in the affected environment, and whether their own records show access or changes during the relevant period. Fireworks’ wording—“a limited set of accounts”—does not provide a count or identify which customers were affected, so customers may need direct confirmation from the company.
The potential exposure is not limited to billing records. A dataset may contain customer material, while model weights can embody work produced by fine-tuning on proprietary information. Fireworks describes the affected weights as incomplete, but has not defined the term or said what those files could reveal. Teams should ask whether their materials were in scope and what steps, if any, are recommended for their specific account.
For security and procurement teams, the incident also raises questions about dependencies beyond a vendor’s own platform. Fireworks attributes one stage to a cloud provider hosting its R&D machines, while saying it has not detected impact to several named products. Customers should distinguish those product-specific statements from a complete account-level assurance: the update does not say which provider was involved, or list all customer accounts and data affected.
Popular right now
Shark NV752 Rotator Powered Lift-Away TruePet Upright Vacuum, Bordeaux
KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer, Blue Velvet
Samsung 43-Inch M70H Series 4K Mini LED Smart TV (2026)
As an affiliate, we earn on qualifying purchases.
From October Disclosure to Update
Fireworks says it identified the initial activity on October 6 and began containment. It began notifying affected customers on October 7, then published its first security bulletin on October 9 after rotating employee API keys. The company says it revoked more user API keys and rotated service-account tokens on October 10.
A separate cloud-provider incident was reported to Fireworks on October 23. The company says the unauthorized workload ran the next day and was detected that day. Fireworks’ November 9 update came about a month after its first public bulletin and added categories of potentially affected information not included in the initial account.
Notion says Fireworks contacted it on November 2 about a dataset initially described as synthetic test data. On November 6, Fireworks told Notion malicious actors may have accessed potential customer data, and the dataset was confirmed on November 8. Notion then disclosed that its AI Insights transcripts were in the dataset and said fewer than 1% of workspaces were potentially affected.
“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 were affected, which accounts were involved, or which cloud provider hosted the R&D machines. It also has not explained how attackers first obtained the internal credentials involved in the initial activity. The company’s description of some models as incomplete does not clarify whether the files were partial copies, unfinished training runs or another form of data.
The available statements do not establish what attackers did with information they accessed, whether it was copied or shared, or whether any customer data has appeared elsewhere. Notion says its transcripts may have been in an accessed dataset, but the advisory does not establish that every transcript was viewed. Fireworks says its investigation is ongoing, so the reported scope may change.
Fireworks also says it has not detected impact to Inference, Training SDK and Nexus. That is a statement about the company’s findings so far, not evidence that all services or all customer data were unaffected. Customers should ask Fireworks for a direct, account-specific assessment rather than infer their status from the general update.
Customer Reviews and Forensics Continue
Fireworks says its investigation is continuing with support from Unit 42 and that it has notified law enforcement. The company has also disabled self-service dataset uploads and downloads used by Batch Inference and dataset-based training while it investigates. It has not provided a date for restoring those functions or for a final report.
Customer teams can ask Fireworks whether their account, evaluation jobs, datasets, model files, environment variables or billing records were accessed; what time period and systems are in scope; and whether any customer action is advised. They can also ask how the company will notify customers if its findings change and what information it can provide about the third-party provider and remediation.
Internally, security, engineering, finance and privacy staff should compare vendor notices with their own logs and data inventories, then follow their established incident-response procedures if exposed credentials or sensitive records may be involved. For organizations using Notion AI Insights, the relevant follow-up is whether their workspace falls within Notion’s potentially affected group and what transcript data was present. The next public milestone is a further update from Fireworks, but no timetable has been announced.
Key Questions
What did Fireworks AI say was affected?
Fireworks says the incident affected account and billing information and, for a limited set of accounts, secrets, datasets and incomplete models. It also lists resource metadata, billing and shipping addresses, invoice and payment history, account balances and credit usage among the information accessible through the affected application.
Were payment card numbers exposed?
Fireworks says it does not store payment card numbers and that they were not affected. This is the company’s statement in its November 9 update.
Was Notion AI data involved?
Notion said chat transcripts from its AI Insights feature were in a dataset attackers may have accessed. It said fewer than 1% of workspaces were potentially affected; that does not establish that every transcript in the dataset was read.
What should Fireworks customers ask their teams to check?
Teams should establish whether they used affected evaluation jobs or stored relevant secrets, datasets or model files with Fireworks, and seek an account-specific assessment from the company. They can compare that response with internal records and follow their incident-response procedures if sensitive information or credentials may be involved.
Has Fireworks said who was responsible?
The supplied update does not name a threat group or the third-party cloud provider. Fireworks says its investigation is ongoing, and the available information does not explain how the initial internal credentials were obtained or what happened to accessed data.
Source: ThorstenMeyerAI.com
