Ants collaborate with Tsinghua University to open source ClawAegis.

On April 1st, Ant Group's AI security laboratory collaborated with Tsinghua University to open-source ClawAegis. According to reports, this is a "native security immune system" designed for OpenClaw-like autonomous intelligent agents, which can real-time identify malicious command injection and intent tampering, prevent models from being "hypnotized," dynamically intervene when the intelligent agent is about to perform sensitive operations, perform permission verification and behavioral audit, and even provide a fuse mechanism for dead loops or resource abuse to ensure the stability of the host machine.