AI Security AI安全 3h ago Updated 1h ago 更新于 1小时前 45

PEEP Turns Chrome and Edge Into Post-Compromise Backdoors for Host Command Execution PEEP将Chrome和Edge变为后渗透阶段的主机命令执行后门

PEEP is a Chromium-based post-compromise toolkit that masquerades as a "Smart Bookmarks" extension to turn Chrome and Edge browsers into persistent backdoors for host-level command execution It bypasses Web Store checks and user prompts by forging Chromium's Secure Preferences integrity values and leveraging native messaging to escape the browser sandbox Built on the open-source RedExt framework, PEEP adds dedicated installation routines, a native host bridge (nm_host.exe), heartbeat telemetry, PEEP是一个Chromium-based后渗透工具包,伪装成书签扩展,可将Chrome/Edge转化为持久化后门 通过伪造Secure Preferences完整性值绕过Web Store检查,利用原生消息传递桥接实现主机级命令执行 基于开源框架RedExt扩展,具备命令执行、凭据窃取、会话劫持和网页篡改能力 使用多个C2端点轮询,每30秒通过明文HTTP通信,窃取浏览历史、Cookie和会话数据 源代码中的中文痕迹指向中文威胁行为者,可能利用"授权CTF"框架降低AI工具安全护栏

65
Hot 热度
70
Quality 质量
55
Impact 影响力

Analysis 深度分析

TL;DR

  • PEEP is a Chromium-based post-compromise toolkit that masquerades as a "Smart Bookmarks" extension to turn Chrome and Edge browsers into persistent backdoors for host-level command execution
  • It bypasses Web Store checks and user prompts by forging Chromium's Secure Preferences integrity values and leveraging native messaging to escape the browser sandbox
  • Built on the open-source RedExt framework, PEEP adds dedicated installation routines, a native host bridge (nm_host.exe), heartbeat telemetry, and an update channel
  • The malware exfiltrates browsing history, session cookies, active-tab metadata, and credentials while executing arbitrary shell commands, managing files, and hijacking sessions via a C2 server polling every 30 seconds over plaintext HTTP
  • Evidence suggests a Chinese-speaking threat actor may have used "Authorized CTF" framing to bypass AI safety guardrails during malware development

Why It Matters

PEEP demonstrates a sophisticated evasion technique where the browser itself becomes the persistence mechanism and execution pivot, bypassing traditional endpoint detection by running malicious logic inside a signed browser process. This represents a growing class of browser-based post-exploitation tools that blur the line between legitimate browser functionality and malware, challenging conventional security monitoring approaches.

Technical Details

  • Installation & Persistence: PEEP uses three PowerShell scripts (install_silent.ps1, patch_secure_prefs.ps1, force_enable.ps1) to enable Developer Mode, sideload the extension, forge Secure Preferences integrity values, and ensure auto-enable on browser launch. It also leverages ExtensionInstallForcelist and ExtensionSettings policies for enterprise force-install scenarios, with a Linux variant using patch_secure_prefs_linux.py.
  • Native Messaging Bridge: The auxiliary executable nm_host.exe (registered as com.peep.lab) acts as a native messaging host that bridges browser commands to OS-level operations, enabling shell command execution, file management, process discovery, and service enumeration while operating in the user context.
  • C2 Communication: The extension polls C2 servers (206.237.30[.]232 or xfjcc[.]fun) every 30 seconds over plaintext HTTP, using endpoints including /api/commands for tasking, /api/exfil for auto-collected data, /api/register for infection registration, /api/agents//heartbeat for telemetry, and /api/extension_update/ for self-updates.
  • Data Exfiltration & Capabilities: Harvests browsing history, active-tab metadata, session cookies, public IP, locale, and timezone. Browser-based commands (screenshots, clipboard access, JavaScript injection) run locally via content.js embedded across all pages, while OS-level tasks are routed through the native host bridge.
  • Evasion Techniques: Runs inside the signed Chrome/Edge process to bypass detection of new or unsigned binaries. Bypasses Web Store verification by manipulating Secure Preferences and using sideloading tricks. The /health endpoint exposes internal system status without authentication.

Industry Insight

  • Security teams should prioritize monitoring for anomalous native messaging host registrations and unexpected modifications to Chromium Secure Preferences files, as these are key indicators of PEEP-style infections that evade traditional binary-focused detection.
  • The use of AI tools with "Authorized CTF" framing to lower safety guardrails during malware development highlights the need for enhanced AI usage monitoring and the implementation of stricter guardrails around code-generation tools in organizational environments.
  • Organizations should audit Chrome/Edge extension policies, particularly ExtensionInstallForcelist and ExtensionSettings, and implement application control policies that restrict unauthorized native messaging hosts to prevent browser-based persistence mechanisms from establishing OS-level access.

TL;DR

  • PEEP是一个Chromium-based后渗透工具包,伪装成书签扩展,可将Chrome/Edge转化为持久化后门
  • 通过伪造Secure Preferences完整性值绕过Web Store检查,利用原生消息传递桥接实现主机级命令执行
  • 基于开源框架RedExt扩展,具备命令执行、凭据窃取、会话劫持和网页篡改能力
  • 使用多个C2端点轮询,每30秒通过明文HTTP通信,窃取浏览历史、Cookie和会话数据
  • 源代码中的中文痕迹指向中文威胁行为者,可能利用"授权CTF"框架降低AI工具安全护栏

为什么值得看

这篇文章揭示了现代浏览器安全的新威胁模式:攻击者利用合法浏览器进程作为端点枢纽,绕过传统检测机制。对于AI从业者和安全研究人员来说,理解这种"合法进程滥用"技术对开发更有效的检测方案至关重要。

技术解析

  • PEEP的核心架构基于Native Messaging Host机制,通过nm_host.exe实现浏览器与操作系统的通信桥接,将浏览器扩展从数据窃取工具升级为远程访问工具
  • 持久化技术包括:伪造Secure Preferences文件、利用ExtensionInstallForcelist策略、侧载技巧、ScriptCache回退机制,以及三个PowerShell脚本(install_silent.ps1、patch_secure_prefs.ps1、force_enable.ps1)
  • C2通信采用明文HTTP协议,每30秒轮询命令,使用多个端点:/api/commands、/api/register、/api/agents//heartbeat、/api/exfil等,C2服务器为206.237.30.232或xfjcc.fun
  • 检测规避策略:利用已签名的浏览器进程执行恶意逻辑,绕过对新建或未签名二进制文件的检测;通过修改Secure Preferences确保扩展在浏览器启动时自动启用
  • 跨平台支持:除Windows外,还存在Linux版本的patch_secure_prefs_linux.py脚本,表明威胁行为者正在扩展攻击范围

行业启示

  • 浏览器安全边界需要重新评估:传统上认为浏览器沙箱能有效隔离恶意代码,但PEEP证明通过Native Messaging等合法机制可以突破这一边界,安全厂商需要重新审视浏览器扩展的权限模型
  • AI辅助恶意软件开发的监管挑战:威胁行为者可能利用"授权CTF"等框架降低AI工具的安全护栏,这提示我们需要在AI安全策略中考虑恶意使用场景,并加强代码生成工具的使用监控
  • 企业端点防护策略需要调整:鉴于PEEP依赖先期入侵,企业应加强初始访问控制,同时部署行为检测而非仅依赖签名检测,特别是监控浏览器扩展的异常行为和Native Messaging通信

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Security 安全 Research 科学研究