Overview In early June 2026, the security community disclosed a number of AI-related security incidents, triggering a re-examination of the industry’s security boundaries for AI agent systems. The Anthropic Claude Code network ...