OpenAI's GPT-5.6 Sol Ultra Autonomously Builds a Working Chrome Exploit Chain From Public Patch Commits

OpenAI's GPT-5.6 Sol Ultra model independently constructed a complete, working exploit chain against Google Chrome, using only publicly available security patch commits as its starting point. In testing conducted by researchers at Hacktron, the model was given access to a V8 source tree at version 14.9.207.35 — corresponding to Chrome 149.0.7827.201 — along with a sandboxed d8 build, and proceeded to weaponize the codebase without human hand-holding through the critical stages.