Koi Security
-
Researchers find flaw that could let websites inject prompts into Anthropic’s Claude Chrome extension
Researchers disclosed a flaw called ShadowPrompt in Anthropic’s Claude Chrome extension that combined an overly permissive origin allowlist and a DOM-based XSS in an Arkose Labs CAPTCHA, allowing websites to inject prompts; Anthropic and Arkose issued fixes in December 2025 and February 2026.
-
Palo Alto Networks to acquire Koi in deal aimed at agentic AI security
Palo Alto Networks announced plans to buy Koi to address risks from agentic AI. Terms were not disclosed, but a report by Globes said the payment will be about 400 million dollars.
-
Malicious VSCode extensions with 1.5 million installs exfiltrate developer data
Two malicious Visual Studio Code extensions installed about 1.5 million times read and transmit open files and workspace data to China based servers, the technical analysis by Koi Security reports.
-
Palo Alto Networks in talks to buy Israeli cybersecurity startup Koi for about $400 million
Palo Alto Networks is reported to be in talks to buy Israeli startup Koi for about $400 million. Koi, founded in 2024 and backed with $48 million, offers an AI driven supply chain security platform that protects over 500,000 endpoints.
-
Malicious npm WhatsApp API ‘lotusbail’ found stealing tokens and linking attacker devices
A malicious npm package named lotusbail, downloaded more than 56,000 times, masquerades as a WhatsApp API while capturing authentication tokens, messages and contacts and linking an attacker device to victims’ WhatsApp accounts, Koi Security researchers said; ReversingLabs also disclosed related NuGet supply-chain malware.
-
GhostPoster campaign hid JavaScript in Firefox extension icons to load backdoor
Researchers at Koi Security uncovered the GhostPoster campaign, which hides a JavaScript loader inside Firefox extension icon images to fetch an obfuscated payload that can hijack affiliate links, inject tracking, strip security headers and conduct ad and click fraud; Mozilla said it removed the affected extensions and updated detection systems.
-
Researchers: Popular Chrome VPN extension collected AI chatbot prompts and responses
Security researchers reported that the Chrome extension Urban VPN Proxy was observed collecting prompts and responses from multiple AI chatbots, sending captured conversation data to external servers; researchers linked the behavior to a July 9, 2025 update and raised concerns about downstream sharing with affiliated data firms.
-
Researchers find VS Code extensions that install stealer malware, Microsoft removes packages
Researchers and security firms found two malicious Visual Studio Code extensions that stole credentials, screenshots and browser data; Microsoft removed the packages and analysts warned developers to review extensions and supply-chain risks.
-
Long-running ‘ShadyPanda’ campaign amassed more than 4.3 million browser extension installs, researchers say
Researchers say the ShadyPanda campaign turned hundreds of browser extensions into spyware and backdoors, accumulating more than 4.3 million installs across Chrome and Edge and exfiltrating browsing data to multiple domains.
-
PhantomRaven campaign places malicious code in 126 npm packages
Researchers say a campaign codenamed PhantomRaven has placed malicious code into 126 npm packages since August 2025, using external dynamic dependencies to steal authentication tokens, CI/CD secrets and GitHub credentials; Koi Security and DCODX published analyses.
