Redesigning Security Culture for the Agentic Age

The report warns that the emergence of agentic AI in the workplace breaks traditional security culture by operating under different patterns of motivation, agency, and normativity. It proposes treating agentic systems like high-risk professions by enforcing immutable observability (logs), rigid structural guardrails, and human-led accountability, and calls for maintaining a minimum viable human workforce with the judgment to oversee and intervene with autonomous agents.