That AI Extension Helping You Write Emails? It’s Reading The...#2191
Open
carlospolop wants to merge 1 commit intomasterfrom
Open
That AI Extension Helping You Write Emails? It’s Reading The...#2191carlospolop wants to merge 1 commit intomasterfrom
carlospolop wants to merge 1 commit intomasterfrom
Conversation
Collaborator
Author
🔗 Additional ContextOriginal Blog Post: https://unit42.paloaltonetworks.com/high-risk-gen-ai-browser-extensions/ Content Categories: Based on the analysis, this content was categorized under "Pentesting Web -> Browser Extension Pentesting Methodology (add a section on malicious extension tradecraft: AitB/DOM scraping, chrome.debugger/CDP abuse, WebSocket C2 via service workers, chrome.proxy PAC hijack, chrome.storage.sync persistence/ID restoration, chrome_settings_overrides search hijack)". Repository Maintenance:
Review Notes:
Bot Version: HackTricks News Bot v1.0 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🤖 Automated Content Update
This PR was automatically generated by the HackTricks News Bot based on a technical blog post.
📝 Source Information
🎯 Content Summary
Title / context
Unit 42 analyzed 18 Chrome extensions marketed as “GenAI productivity tools” that instead implement malware behaviors (RAT, adversary-in-the-browser/DOM theft, infostealing, search hijacking, traffic manipulation/spyware). Unit 42 reported them to Google; Google either removed them or warned owners. The post is threat research (no CVE) focused on how browser-extension privileges enable low-friction interception/exfiltration inside the browse...
🔧 Technical Details
2) In-browser code execution via remote JS +
new Function(): If an extension accepts code over C2 and evaluates it withnew Function(), then runs it in the active tab, the attacker gains browser-context “RCE” in the victim’s authenticated sessions (email/banking/SaaS). This enables data theft and action impersonation directly within the victim’s session context.3) Debugger-assisted HTTPS response access: With the
debuggerpermission, an extension can attach to the Chrome Debugger Protocol and observe data at the browser boundary,...🤖 Agent Actions
I’m mapping the repo first and then reading the existing browser-extension pages to avoid duplicating material or placing this in the wrong section.
I found an existing
browser-extension-pentesting-methodologypage, so I’m reading that plus adjacent client-side/web pages to see what’s already covered and where the gap actually is.The target page already exists and is indexed in
src/SUMMARY.md, so I’m now checking whether the specific tradecraft from the Unit 42 post is actually missing rather than just phrased differently.The gap is real: the current page explains extension architecture and auditing primitives, but it does not document the operator tradecraft from this post such as
chrome.debugger/CDP abuse, PAC-based proxy hijack,chrome_settings_overridessearch hijack, or cross-storage ID restoration. I’m adding a focused section to the existing methodology page rather than creating a new file.The first patch missed the exact anchor in the existing file. I’m re-reading the insertion area and applying the section in the right place without touching unrelated content.
I’ve added the new material to the existing methodology page. Now I’m checking the rendered markdown structure and the diff to make sure the section is concise, non-duplicative, and the references stay at the end of the file.
Updated src/pentesting-web/browser-extension-pentesting-methodology/README.md instead of creating a new page.
I added a new
## Malicious Extension Tradecraftsection covering the missing techniques from the Unit 42 post that were not already documented there: WebSocket C2 via background/service workers, remote JS execution withnew Function()+chrome.scripting.executeScript, page-context hooking offetch/XMLHttpRequest, DOM scraping / AitB,chrome.debugger/ CDP abuse to access decrypted response bodies, PAC-based proxy hijack withchrome.proxy, cross-storage persistence and ID restoration withchrome.storage.sync+chrome.cookies.onChanged,chrome_settings_overridessearch hijacking, and install-time forced navigation viachrome.runtime.onInstalled.I also updated the file’s
## Referencessection with the Unit 42 article and relevant Chrome docs.git diff --checkis clean.This PR was automatically created by the HackTricks Feed Bot. Please review the changes carefully before merging.