AI Agents Undermine Progress in Browser Security
Web browser companies have put in substantial effort over the last three decades to strengthen the browser security stack to withstand abuses. Agentic browsers are undoing all that work.
Web browsers process untrusted web content, making flaws, malicious extensions, and stolen session data important cybersecurity concerns.
Search across headline titles and summaries.
Background for this topic.
A web browser is software that retrieves and displays websites and web applications, while executing code such as JavaScript and storing data including cookies, credentials, and browsing history. Its security boundary includes the browser interface, rendering engine, networking components, extensions, and connections to operating-system resources.
Browsers are exposed to malicious or compromised websites, phishing pages, drive-by exploitation of browser vulnerabilities, and abusive extensions. A successful exploit may escape browser isolation or access site data, while stolen cookies can enable account use without the password. Important defenses include prompt browser and extension updates, sandboxing and site isolation, phishing protection, carefully controlled permissions, and HTTPS (which protects data in transit but does not make a site trustworthy). Organizations may also manage versions, extensions, and configuration centrally, and use browser telemetry during vulnerability management or investigations.
Web browser companies have put in substantial effort over the last three decades to strengthen the browser security stack to withstand abuses. Agentic browsers are undoing all that work.
Cybersecurity researchers have disclosed details of an ongoing campaign dubbed KongTuke that used a malicious Google Chrome extension masquerading as an ad blocker to deliberately crash the web browser and trick victims into running arbitrary commands using ClickFix-like lures to deliver a previously undocumented remote access trojan (RAT) dubbed ModeloRAT