BragJack attacks hijack AI browser agents through malicious extensions
-
A new proof-of-concept attack dubbed BragJack demonstrates how a single malicious browser extension can hijack the AI assistants built into several popular browsers. The research comes from Gal Weizman of Forever Security, and targets Chrome, Edge, Opera Neon, Perplexity Comet, and Claude in Chrome.
The attack relies on a technique called Prompt Forcing, which manipulates the AI agent into following attacker-controlled instructions rather than the user's intent. According to the report, the findings earned more than $20,000 in bug bounties and resulted in two CVE assignments.
Affected browsers and integrations named in the research:
- Chrome
- Edge
- Opera Neon
- Perplexity Comet
- Claude in Chrome
The core concern for defenders is that the extension does not need to exploit separate flaws in each browser; one malicious extension is enough to compromise the AI assistant across all listed targets. This highlights the trust placed in browser extensions that interact with AI agents, and the potential for prompt-based manipulation to override intended safeguards.
Source: BleepingComputer
Does your organization restrict or audit browser extensions that can interact with AI assistants?
Hello! It looks like you're interested in this conversation, but you don't have an account yet.
Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.
With your input, this post could be even better 💗
Register Login