Offensive Embedded Exploitation : Getting hands dirty with IOT/Embedded Device Security Testing

By Kaustubh Padwad on 18 Sep 2020 @ C0c0n
πŸ“Ή Video πŸ”— Link
#red-teaming #security-assessment #iot-device-management #iot-pentesting #embedded-systems #firmware-analysis #dynamic-binary-instrumentation
Focus Areas: πŸ”§ Hardware Security , 🏭 Industrial Control Systems Security , πŸ“‘ IoT Security , 🎯 Penetration Testing , πŸ”¬ Reverse Engineering , πŸ” Vulnerability Management

Presentation Material

AI Generated Summary

Here is a summarized version of the content:

The speaker, an expert in exploiting devices, shares their experience with crashing a device using HTTP reflection. They created a list of password attempts (2, 5, 10, 20, 50, 100, 200 times) and loaded it into a payload to test the device’s response. The device crashed multiple times, and the speaker analyzed the log files to understand what happened in the background.

The speaker emphasizes the importance of testing each request with every parameter to avoid surprises. They used Metasploit (MSF) to generate a payload and identify the offsets at which the device crashes. To develop a successful exploit, they recommend:

  1. Obtaining a root shell
  2. Finding the process ID of the running process
  3. Installing a GDB server
  4. Attaching the process to GDB
  5. Regenerating the crash to gain control over registers and written addresses

The speaker demonstrates an exploit they developed for an iPhone, using a buffer overflow attack with a specific username and password. They highlight the importance of hosting registers and providing padding to own the device without interaction.

Finally, the speaker invites questions and encourages further research, offering their contact information for those interested in learning more.

Disclaimer: This summary was auto-generated from the video transcript using AI and may contain inaccuracies. It is intended as a quick overview β€” always refer to the original talk for authoritative content. Learn more about our AI experiments.