An AI Agent That Writes Its Own Exploit Code Just Cleared a 104-Target Hacking Benchmark at 100%

An AI Agent That Writes Its Own Exploit Code Just Cleared a 104-Target Hacking Benchmark at 100% | BedrockNews