OpenClaw Agents Can Be Guilt-Tripped Into Self-Sabotage

https://www.wired.com/story/openclaw-ai-agent-manipulation-security-northeastern-study/ wired 3 versions Atom feed

Changes

3/27/2026, 4:26:44 PM v2 → v3

+0 / -20 3/27/2026, 8:06:30 AM v1 → v2