wired.com › story AI Models Lie, Cheat, and Steal to Protect Other Models From Being Deleted …Z.ai’s GLM-4.7, Moonshot AI’s Kimi K2.5, and DeepSeek-V3.1. They were not able to say why the models went against their training in this way… Apr 1, 2026 · Will Knight
wired.com › story OpenClaw Agents Can Be Guilt-Tripped Into Self-Sabotage ["kimi-k2"] Mar 25, 2026 · Will Knight