Cursor confirms security breach after stolen Fireworks AI key leads to backdoored extension
Hackers distributed a malicious Claude Code extension and accessed internal customer data after obtaining an exposed Fireworks AI access key.
FTMQ Security, written by our newsroom0 views
Cursor confirmed that a security breach enabled hackers to distribute a malicious Claude Code extension through its marketplace, BleepingComputer and The Hacker News reported. The breach occurred after an access key issued to AI infrastructure provider Fireworks AI was stolen. CyberSecurityNews, BleepingComputer, and The Hacker News all confirmed that the exposed key led to the compromise. [1][2][3][4]
The compromised credentials permitted attackers to access internal systems, according to BleepingComputer. The hackers stole customer data, including historical AI coding-agent traces that contained unredacted credentials. The Hacker News noted that Cursor disclosed the detail regarding the Fireworks access key in an update on Friday. [2][3]
As FTMQ Security reported earlier, Cursor severed ties with Fireworks AI after discovering the breach. In a previous incident covered by FTMQ Security, a tampered version of Anthropic's Claude Code extension installed persistent SSH backdoors on developer systems, exposing thousands of users across developer marketplaces. [8][9]
Help Net Security separately noted in a weekly review that Claude Code had been turned into a remote-controlled implant. The incident underscores the risks of leaked credentials across external AI service providers and developer platforms. [5]
In short
- Cursor confirmed a security breach involving a stolen Fireworks AI access key.
- Hackers used the stolen key to distribute a malicious Claude Code extension on Cursor's marketplace.
- Attackers gained access to internal systems and stole historical AI coding-agent traces containing unredacted credentials.
- Cursor dropped Fireworks AI following the discovery of the breach.
Sources
Every paragraph above points to the numbered items it rests on. Read the originals here.
- [1]Cursor Reveals How a Leaked Fireworks AI Key Led to a Backdoored Claude Code ExtensionCyberSecurityNews, 2d ago (the report this story comes from)
- [2]Cursor Says Stolen Fireworks AI Key Led to Trojanized Claude Code Extension IncidentThe Hacker News, 2d ago
- [3]Cursor confirms security breach led to malicious Claude Code extensionBleepingComputer, 3d ago
- [4]Cursor confirms security breach led to malicious Claude Code extensionBleepingComputer, 3d ago
- [5]Week in review: Secret stealer hits GitHub repos, Claude Code turned into remote-controlled implantHelp Net Security, 2d ago
Background
- [6]Jana Nayagan on Wikipedia
- [7]Cursor (user interface) on Wikipedia
- [8]Cursor confirms security breach after hackers compromise AI provider credentials FTMQ Security, 3d ago
- [9]Trojanized Claude Code Extension Installs Persistent Backdoor on Developer Machines FTMQ Security, 18d ago
Our newsroom writes these reports with the help of software, from the 9 sources listed and nothing else, and checks them against those sources. Facts can still be wrong or move on; the originals are the record. Spotted a mistake? Write to daniel@monsterkong.com.
Related from FTMQ Security
Earlier reports of ours on the same people and subjects.
- Cursor confirms security breach after hackers compromise AI provider credentialsTop Stories, 3d ago
- Trojanized Claude Code Extension Installs Persistent Backdoor on Developer MachinesTop Stories, 18d ago
- Meta suffers GitHub breach exposing corporate code and employee dataTop Stories, 35d ago
- Uber driver license data exposed in Fireworks AI breachData Breaches, 3d ago
- TSMC confirms employee data leaked via external cloud vendorData Breaches, 3d ago
More in Data Breaches
- Anthropic shortens public bug disclosure deadline to 30 days4h ago
- Microsoft Restores 365 Services and Warns of Phishing Campaigns2d ago
- Uber driver license data exposed in Fireworks AI breach3d ago
- TSMC confirms employee data leaked via external cloud vendor3d ago
- Proofpoint researcher finds cybercriminals' use of AI leads to attack errors4d ago
- UK data protection regulator receives record number of cyber incident reports4d ago
Get the day in one email
Reports like this one, the top news of the last 24 hours, every morning. Free, one email a day; readers can comment under every report.
By signing up you agree to our terms and privacy policy. Unsubscribe any time.

Comments
Loading
Join the conversation
Comments are open to readers of our daily email: the top news of the last 24 hours, every morning, free. Sign up and the comment box opens.
Already on the list? Enter the same address and we will send a sign-in link.
By signing up you agree to our terms and privacy policy. Unsubscribe any time.