For CISOs adopting agentic AI, the most practical first step is to frame it as an insider risk problem. This involves assigning agents persistent identities (like Slack or email accounts) and applying rigorous access control and privilege management, similar to onboarding a human employee.

Since credential theft is rampant, authenticating users at login is insufficient. A modern security approach must assume breach and instead focus on anomalous behavior. It should grant access dynamically and "just-in-time" for specific tasks, revoking rights immediately after.

Jamie Dimon uses travel and site visits as a primary tool for uncovering operational flaws. He returns with a detailed list of questions and required actions, creating a relentless feedback loop that forces accountability and prevents complacency among senior leaders.

Jamie Dimon rejects conventional risk models that test for modest downturns (e.g., a 10% market drop). He forces his team to model for catastrophic, 'worst ever' events to truly understand and prepare for tail risk, which 'undresses how much risk people are taking.'

Instead of relying on flawed AI guardrails, focus on traditional security practices. This includes strict permissioning (ensuring an AI agent can't do more than necessary) and containerizing processes (like running AI-generated code in a sandbox) to limit potential damage from a compromised AI.

Blankfein believes the biggest technological threat isn't a sophisticated cyberattack but a simple human mistake amplified by technological leverage. He warns that adding more layers of checks can create complacency, paradoxically making such an error more likely to slip through.

To balance agility and scale, Jamie Dimon structures teams like Navy SEALs. Small, dedicated groups are fully authorized to complete a mission, preventing bureaucratic drag. However, they use common equipment and platforms, avoiding the chaos of total decentralization.

AI agents can cause damage if compromised via prompt injection. The best security practice is to never grant access to primary, high-stakes accounts (e.g., your main Twitter or financial accounts). Instead, create dedicated, sandboxed accounts for the agent and slowly introduce new permissions as you build trust and safety features improve.

To counteract his own forceful personality and enable candid discussion, Jamie Dimon mandates that his board holds a session without him at every single meeting. The lead director then provides him with direct coaching and feedback, creating a powerful accountability mechanism.