0 views collected around this technical thread.
The article reports on the GeekPwn2018 security competition in Shanghai, highlighting JD security teams' IoT research, explaining how OBD‑2 based car insurance devices can be hijacked, and stressing the urgent need for stronger IoT protection against physical‑layer attacks.
JD Security researchers attended the BRUCON conference to unveil a new fuzz‑testing operation that extends AFL with custom instrumentation, enabling faster testing across multiple CPU architectures and overcoming the limitations of existing fuzzers for closed‑source and non‑x86 systems.
The article explains how the black‑box nature of deep learning creates security risks such as adversarial attacks, describes real‑world examples in autonomous driving and medical imaging, and showcases JD Security's explainable AI system that demystifies model decisions to improve AI safety and industry adoption.