Notion confirms AI chat exposure in Fireworks supplier breach

Notion's November 10 advisory, updated November 11, says a Fireworks AI incident exposed chat transcripts used for its Insights feature, affecting fewer than 1% of workspaces. SecurityWeek reported the case November 12.

Notion says most affected transcripts lacked author and workspace labels and tool-call results. That does not strip identifying details from the conversation itself. It disabled Insights, stopped serving models through Fireworks and revoked exposed Notion API keys found in messages on November 11.

Fireworks' November 9 bulletin traces one intrusion to October 24 through a cloud supplier hosting development machines. This is not a November 12 attack. Its investigation continues; it reports no detected impact to Inference, Training SDK or Nexus.

Notified customers should inspect the affected data and rotate credentials they pasted into those chats. Notion says customers it has not contacted need take no action right now, while investigation continues.

Official advisories: https://www.notion.com/en-gb/security/advisories/security-incident-involving-a-third-party-ai-vendor https://fireworks.ai/security/bulletins/october-2026-security-incident-update Reporting: https://www.securityweek.com/notion-customer-data-compromised-in-fireworks-ai-hack/

Original source published: SecurityWeek, Nov 12, 2026, 19:08 Beijing time / 6:08 AM EST. Notion dates its original advisory Nov 10 and latest update Nov 11; Fireworks dates its update Nov 9. Those official pages provide no exact time or timezone, so no conversion is asserted.