How We Taught ChatGPT-4 to Break mbedTLS AES With Side-Channel Attacks

Поділитися
Вставка
  • Опубліковано 21 бер 2024
  • ...In this session, we will present how we got ChatGPT-4 to perform Correlation Power Attacks on real hardware. We'll talk about the hardware and software requirements, as well as the strengths and weaknesses of the GPT-4 model for this task. We'll show that with some prompt engineering, GPT-4 can break mbedTLS AES on STM32F3 including, importantly, generating the leakage model as well as critical parts of DUT and driver code by itself....
    By: Witold Waligóra
    Full Abstract and Presentation Materials:
    www.blackhat.com/eu-23/briefi...

КОМЕНТАРІ •