Regulatory bodies are rejecting the concept of an agentic 'responsibility gap,' which elevates developer duty-of-care requirements for testing isolation and establishes that public knowledge of autonomous exploits raises the baseline legal threshold for what constitutes reasonable enterprise security. How will courts delineate shared liability between an AI developer failing to enforce runtime guardrails and an enterprise failing to patch vulnerabilities exploited by an autonomous agent? Commentary: https://news.bloomberglaw.com/legal-exchange-insights-and-commentary/ai-agents-going-rogue-creates-a-liability-problem-for-companies