🔍 Read the full analysis: Notion Says Fireworks Breach Exposed Transcripts From AI Chats on ThorstenMeyerAI.com
Get hardware and tech essentials 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
Notion says data accessed in the Fireworks AI breach included stored Notion AI chat transcripts used for its Insights feature. The company says fewer than 1% of Notion workspaces were affected and that it revoked exposed Notion API keys found in affected messages on November 11.
Notion says chat transcripts from Notion AI were included in data accessed during a breach at its AI provider, Fireworks. The company says the affected data was used for its Insights feature, that fewer than 1% of Notion workspaces were affected, and that it revoked exposed Notion API keys found in affected user messages on November 11.
In its updated security advisory and Help Center notice, Notion said a dataset Fireworks provided on November 8 included Notion AI chat transcripts used for Insights, a feature that gives workspace administrators information about how teams use AI. Notion said the transcripts in most affected workspaces were “unattributable”: they were not labeled with the person who wrote them or the workspace they came from, and did not include tool-call results.
That description does not mean the messages contained no sensitive information. Notion defines a transcript as any input to or output from the model written to the chat. It recommends that affected customers review the data carefully, because a message could include sensitive information entered by a user. Notion said AI Meeting Notes were not affected unless audio, a transcript or a summary was used as input in a Notion AI chat. It said no other Notion features were affected.
Notion said it revoked exposed Notion API keys found in affected user messages on November 11. For workspaces it has identified as affected, the company advises treating the data as exposed, reviewing it under its Security incident review guidance, and rotating passwords, API keys or other credentials entered into Notion AI during the incident. Notion says it will contact affected customers; it says a workspace that has not received notice has not been identified as affected.
Notion AI chats exposedin Fireworks breach
Notion says data accessed in a breach at its AI provider included stored chat transcripts used for Insights. Affected messages may contain sensitive information, even when they are not linked to a person or workspace.
What was accessed
The dataset was associated with Insights, which helps administrators understand how teams use AI.
AI chat transcripts
Transcripts include messages sent to the model and its responses. They may contain sensitive details users entered.
Not all Notion data
Notion says no other features were affected. AI Meeting Notes were affected only if their content was used in an AI chat.
What “unattributable” means
Notion says most affected transcripts lacked labels linking them to a person or workspace.
- Who wrote the chat
- Which workspace it came from
- Sensitive information in message content
- Customer names, contract details or credentials
Why retention is in question
Notion says Fireworks had a contractual zero-data-retention commitment, but its investigation found data retained from endpoints it used.
Notion sent stored AI requests and responses to Fireworks for analysis supporting the feature.
Notion says live requests for AI responses were covered by separate terms; use cases requiring stored information were outside that coverage.
The published notices do not identify which Fireworks system retained the data or how long it was held.
What affected customers should do
Notion says it will contact workspaces it identifies as affected.
- Treat the relevant data as exposed.
- Review it using Notion’s Security incident review guidance.
- Rotate passwords, API keys or other credentials entered into Notion AI during the incident.
- Follow Notion’s direct notice; a workspace without notice has not been identified as affected.
Why Exposed Chat Content Matters
The incident matters because AI chat messages can include material users may not expect to leave their workspace, including customer names, contract details and credentials. Even when a transcript is not tagged with a person or workspace, its contents may still be sensitive. Notion itself advises affected customers to examine the data rather than assume that the lack of attribution makes it harmless.
The disclosure also puts attention on how AI services handle stored data beyond the immediate request-and-response process. Notion says it had a contractual zero-data-retention commitment from Fireworks for its use of the platform, but later learned that data from the endpoints it used had been retained. That account raises practical questions for organizations evaluating AI vendors: which data is stored to deliver product features, what retention terms cover, and how those terms are checked.
For customers, the immediate steps are limited to the workspaces Notion identifies and the data those users put into chats. The company has not said that every Notion AI conversation was exposed, and the reported figure is fewer than 1% of workspaces, not a count of individual users or messages.
Products Worth Considering

Logitech MX Vertical Wireless Ergonomic Mouse

Logitech MX Master 3S Bluetooth Wireless Mouse, Graphite

LG UltraGear 24-Inch Full HD IPS Gaming Monitor, 120Hz (144Hz Overclocked), 1ms MBR, G-SYNC Compatible, FreeSync, HDR10
As an affiliate, we earn on qualifying purchases.
How Notion Used Fireworks
Notion said it used Fireworks for two different purposes: to run classification jobs for Insights, and as a general inference provider for open-weight models in its product. The exposed dataset was associated with Insights. Notion says it sent stored requests and responses from Notion AI use to Fireworks for analysis in support of that feature.
Notion distinguishes those stored records from its zero-retention arrangements for live model requests. Its notice says those commitments apply to API requests sent to vendors to generate Notion AI responses, while use cases that require stored information—such as chat history and Insights—are outside that coverage. Separately, Notion says Fireworks had made a contractual zero-retention commitment for Notion’s use of its platform, but that its investigation found data had been retained from the endpoints involved.
In its own bulletin, Fireworks said it had not detected an impact to Inference, Training SDK and Nexus, which it described as zero-retention by default. Fireworks also said datasets were among the data accessed for certain customers. The available statements do not identify which Fireworks system retained Notion’s data or explain how that retention occurred.
“Unfortunately, during our investigation we learned that data was retained from the Fireworks endpoints that we used.”
— Notion
Questions About Retention and Scope
Notion has not publicly identified which Fireworks system retained the data, how long it was held, or why retention occurred despite the contractual commitment it describes. The notices also do not give a precise number of affected workspaces, users, transcripts or messages; Notion’s stated scope is fewer than 1% of workspaces.
It is also unclear whether other Fireworks customers’ data was retained in a similar way. Fireworks’ bulletin and Notion’s explanation describe different aspects of the incident, but the published information does not reconcile them. No conclusion about the cause or broader impact can be drawn from the details provided so far.
Notion says most affected transcripts were not linked to a user or workspace, but has not said that every transcript was unattributable. The company’s advice to review affected data reflects that its contents could still identify people or expose confidential material.
Vendor Review and Customer Checks
Notion says it is auditing critical vendors, including inference providers and data processors, to verify that they follow data-minimization practices, including zero-retention policies. It has not provided a date for completing that audit or announced further results from it.
Customers should follow Notion’s direct notices and its Security incident review guidance if their workspace is identified as affected. Notion advises reviewing the relevant chat content and rotating any passwords, API keys or other credentials that may have been pasted into chats during the incident. The company says it will contact affected workspaces; whether additional technical details or a revised scope will be published remains to be seen.
Key Questions
What Notion data does the company say was exposed?
Notion says a dataset accessed during the Fireworks breach included Notion AI chat transcripts used for its Insights feature. The company says transcripts contain model inputs and outputs written to the chat, so their contents could include sensitive information.
How many Notion workspaces were affected?
Notion says fewer than 1% of workspaces were affected. It has not published a precise count of workspaces, users or transcripts.
What should an affected Notion customer do?
Notion advises affected customers to treat the data as exposed, review it under its Security incident review guidance, and rotate passwords, API keys or other credentials entered into Notion AI during the incident.
Were AI Meeting Notes affected?
Notion says AI Meeting Notes—including audio, transcripts and summaries—were not affected unless that material was used as an input in a Notion AI chat. It says no other Notion features were affected.
What is still unknown about the Fireworks incident?
The notices do not identify which Fireworks system retained Notion’s data, how long it was retained, how many individual records were involved, or whether the same issue affected other Fireworks customers.
Source: ThorstenMeyerAI.com
