🔍 Read the full analysis: Notion Links AI Chat Transcript Exposure To Fireworks Breach on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Notion says data accessed in the Fireworks AI breach included stored Notion AI chat transcripts used for its Insights feature. Fewer than 1% of Notion workspaces are identified as affected, and the company says it revoked exposed Notion API keys found in affected messages on November 11. Notion also says Fireworks had promised zero data retention but data was retained; the system involved and retention period remain unclear.
Notion says chat transcripts from its AI service were included in data accessed by attackers during the Fireworks AI breach, expanding its earlier notice about the incident. The company says fewer than 1% of Notion workspaces are affected and that it revoked exposed Notion API keys found in affected user messages on November 11.
According to Notion’s updated security advisory, the exposed dataset included Notion AI chat transcripts used for its Insights feature. Insights gives workspace administrators information about how their teams use AI. Notion says the transcripts were provided to Fireworks for analysis and that, for most affected workspaces, they could not be attributed to a particular person or workspace. It defines “unattributable” as chats lacking labels identifying their author or workspace; it also says the transcripts did not contain tool-call results.
That limitation does not mean the text itself was necessarily harmless. Notion defines a transcript as model input or output written to the chat and advises affected customers to review the data because it may contain sensitive information. The company says AI Meeting Notes—including audio, transcripts and summaries—were not affected unless they were used as input in a Notion AI chat transcript. It says no other Notion features were affected.
Notion’s notice says Fireworks had made a contractual zero-data-retention commitment for Notion’s use of its platform. Notion later learned that data was retained from the Fireworks endpoints it used. On November 11, it said it had revoked exposed Notion API keys found in affected user messages. It advises affected customers to treat the data as exposed, review it through its Security incident review guidance and rotate any passwords, API keys or other credentials entered into Notion AI during the incident.
Notion AI transcripts exposedin Fireworks breach
Notion says attackers accessed stored chat transcripts used for its Insights feature. Affected customers should check for sensitive information and credentials.
What was in the dataset
Notion says the accessed data included stored requests and responses analyzed for Insights.
Inputs and outputs
Transcripts may contain text people entered directly, including sensitive information.
Usually unlabeled
Most affected transcripts could not be linked to a particular person or workspace.
Limited feature impact
Notion says no other features were affected; Meeting Notes were excluded unless used as chat input.
What customers should do
Notion recommends treating affected data as exposed.
Use Notion’s Security incident review guidance to check transcripts for sensitive content.
Change passwords, API keys, or other credentials entered into Notion AI during the incident.
Incident timeline
What remains unclear
- Fireworks retained data despite a contractual zero-retention commitment.
- The affected data came from Notion’s Insights analysis work.
- Which Fireworks system held the data or how long it was retained.
- How many individuals were represented, or whether other customers were affected.
Sensitive Text May Be in Transcripts
The exposure matters because a chat transcript can contain information a person entered directly, not just a summary of how an AI feature was used. Notion specifically warns that affected input or output may contain sensitive data. People and organizations that pasted credentials, customer details or other confidential material into chats may need to check whether those items appear in the affected data and take protective steps.
The incident also puts a boundary around what a zero-retention promise covered in this case. Notion says its own zero-retention commitments for live requests to model providers do not cover all data stored to support features such as chat history and Insights. For Insights, the company says stored requests and responses were sent to Fireworks for analysis under a separate contractual commitment. Notion’s account indicates that data was retained despite that commitment, but does not explain how or why.
Notion says the affected group is fewer than 1% of workspaces, but that percentage does not reveal the number of individual users or the contents of each transcript. For a workspace that is affected, the practical issue is whether a transcript included information that remains sensitive and whether any credentials need to be changed.
Popular right now
Ninja Crispi 4-in-1 Portable Glass Air Fryer, Grey, FN101GY
iRobot Roomba 105 Combo Robot Vacuum & Mop with AutoEmpty Dock
Bose QuietComfort Wireless Over-Ear Noise-Cancelling Headphones, Sandstone
As an affiliate, we earn on qualifying purchases.
How Notion Used Fireworks
Notion says it used Fireworks for two purposes: to run Insights classification jobs and to serve open-weight models within its product. The exposed data described in the updated notice came from the Insights work, where stored Notion AI requests and responses were analyzed. Notion says the incident did not affect other Notion features, subject to its qualification about AI Meeting Notes being used as chat input.
In its published timeline, Notion says Fireworks notified it on November 2 that a small set of data had been accessed by malicious actors. Notion initially identified the material as synthetic testing data. On November 3, it says it disconnected its systems from Fireworks, deleted data from the Fireworks system and stopped communications between the companies’ systems. It also says it halted training runs. Notion updated its advisory on November 10 and 11 as its investigation developed.
Fireworks’ own bulletin, according to the source material, remained marked as last updated November 9. Fireworks said it had not detected an impact to Inference, Training SDK and Nexus, which it described as zero-retention by default, while listing datasets among information accessed for certain customers. The available statements do not identify which Fireworks system retained Notion’s data or explain how the two companies’ descriptions relate.
“Unfortunately, during our investigation we learned that data was retained from the Fireworks endpoints that we used.”
— Notion, in its updated security notice
Retention Details Remain Unknown
Notion has not identified which Fireworks system retained the data, how long it was kept or how many individuals are represented in the affected workspaces. Its notice gives a workspace-level estimate—fewer than 1%—but does not provide a total number of affected users or a detailed breakdown of transcript contents.
The public statements also do not explain why data was retained despite the contractual commitment, or whether the same issue affected other Fireworks customers. Fireworks’ statement that it detected no impact to certain services does not, on the information provided, resolve the question of where Notion’s stored Insights data was held. The available material does not confirm that exposed transcripts were accessed or read by attackers beyond being included in the accessed dataset.
Notion Reviews Vendors and Customer Data
Notion says it is auditing critical vendors, including inference providers and data processors, to check their data-minimization practices and zero-retention policies. The company has not given a completion date for that review. Customers notified by Notion should follow its Security incident review guidance and check affected material for credentials or other sensitive information.
For affected users, Notion recommends rotating passwords, API keys and other credentials entered into Notion AI during the incident. Further clarification from Notion or Fireworks could establish which systems held the data, how long retention lasted and whether the issue extended to other customers. No additional investigation findings or timetable are specified in the source material.
Key Questions
What Notion data was exposed in the Fireworks breach?
Notion says the dataset included Notion AI chat transcripts used for Insights, containing model input or output written to the chat. The company says the transcripts generally lacked labels identifying the author or workspace, and did not contain tool-call results.
How many Notion workspaces are affected?
Notion puts the affected share at fewer than 1% of workspaces. The notice does not give a count of individual users, and the estimate should not be read as a specific number of people.
Were Notion AI Meeting Notes affected?
Notion says AI Meeting Notes, including audio, transcripts and summaries, were not affected unless used as input in a Notion AI chat transcript. It says no other Notion features were affected.
What should affected users do?
Notion advises affected customers to treat the data as exposed, review it under its Security incident guidance and rotate credentials—including passwords and API keys—if they entered them into Notion AI during the incident.
Why was data retained if Fireworks promised zero retention?
Notion says Fireworks made a contractual zero-data-retention commitment, but that its investigation found data had been retained from the endpoints it used. The companies have not publicly specified which system retained the data, why it happened or for how long.
Source: ThorstenMeyerAI.com
