BragJack attacks hijack AI browser agents through malicious extensions
What happened
Security researcher Gal Weizman disclosed BragJack, a proof-of-concept attack that uses a single malicious browser extension to hijack AI assistants in five Chromium-based browsers. By exploiting Chromium's declarativeNetRequest functionality to weaken security headers and redirect resources, the extension executes code within the AI agent's context, granting it access to local files, camera, and microphone. Weizman demonstrated that a malicious extension could force Perplexity Comet's agent to summarize the victim's emails and send the results to another address.
Why it ranks #1
Gal Weizman disclosed BragJack, a proof-of-concept using a malicious extension to hijack AI agents in five Chromium browsers by exploiting declarativeNetRequest to weaken security headers and redirect resources.
Who should care
CISOs and security leaders, SOC and incident response teams
What to do
Keep browsers updated, remove unrecognized extensions, and treat broad "read and change all your data" permission prompts with caution.
- Impact
- high
- Urgency
- near-term
- Confidence
- high
- Scope
- enterprise
- Status
- proof of concept