
What happened
Fireworks AI said attackers breached its cloud provider to reach a development system, taking customers' billing information and some datasets. Notion, which uses Fireworks, confirmed hackers stole chat transcripts from Notion AI's "Insights" feature.
Why it matters
Notion said the stolen data was largely unattributable to specific customers, showing how a breach at one provider can spill over into another company's users.
WHO IT HITSDevelopers and companies that build on Fireworks AI's platform face exposure of billing data and datasets, while Notion AI users who used the Insights feature may have had their chat inputs and outputs stolen.
Summaries like this, every morning.
The incident at Fireworks AI highlights a structural dependency in the AI toolchain: companies like Notion build their own AI features on top of third-party model-access platforms, so a compromise at the provider can propagate to the provider's customers. Fireworks touts tens of thousands of customers worldwide, including legal tech company Harvey and AI company Cursor, which widens the potential surface of any such spillover.
Notion's response was to sever ties entirely on November 3, deleting the data stored in its systems and stopping all connectivity with Fireworks. That the stolen transcripts were largely unattributable to specific customers complicates notification and remediation, since the company cannot always determine whose data was taken.
Fireworks said it believes the hackers are financially motivated and focused on selling tokens from breached access to large language models, a scheme that lets unauthorized third parties burn through AI model usage without using their own resources. This is the second breach affecting Fireworks in as many months, following a security incident in October.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
The author, writing on day 42 of what he calls the "Black October" wave of AI-discovered security flaws, says…

Australia announced a law forcing frontier AI companies to open models to independent audits, publish safety f…

Gerald Mullally, CEO of Oxford Quantum Circuits, said quantum machines will hit millions of steps before an er…

Jeff Bezos said AI productivity gains could let two-income households live on one paycheck or a three-day week…

Tempura technologies joined NVIDIA Inception and will register its Securiy security AI service, verifying NVID…

Chris Betz published "Cloud CISO Perspectives: Our security forecast for 2027," warning that attackers will po…
