The house sitter problem: Why AI agents still have the keys
What happens when access outlives intent? This article uses a house‑sitter analogy to explain how AI agents turn lingering access and expanding scope into governance risk.