To safely use Clawdbot, the host created a dedicated ecosystem for it: a separate user account, a unique email address, and a limited-access password vault. This 'sandboxed identity' approach is a crucial but non-obvious security practice for constraining powerful but unpredictable AI agents.

To manage security risks, treat AI agents like new employees. Provide them with their own isolated environment—separate accounts, scoped API keys, and dedicated hardware. This prevents accidental or malicious access to your personal or sensitive company data.

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.

Simply giving an agent a user account is dangerous. An agent creator is liable for its actions, and the agent has no right to privacy. This requires a new identity and access management (IAM) paradigm, distinct from human user accounts, to manage liability and oversight.

Current agent frameworks create massive security risks because they can't differentiate between a user and the agent acting on their behalf. This results in agents receiving broad, uncontrolled access to production credentials, creating a far more dangerous version of the 'secret sprawl' problem that plagued early cloud adoption.

Current AI tools are in "easy mode" because they operate with the user's direct authentication and permissions. The much harder, yet-to-be-solved problem is "hard mode": autonomous agents that need their own scoped access to enterprise resources without dramatically increasing security risks.

Managing human identities is already complex, but the rise of AI agents communicating with systems will multiply this challenge exponentially. Organizations must prepare for managing thousands of "machine identities" with granular permissions, making robust identity management a critical prerequisite for the AI era.

To address security concerns, powerful AI agents should be provisioned like new human employees. This means running them in a sandboxed environment on a separate machine, with their own dedicated accounts, API keys, and access tokens, rather than on a personal computer.

Todd McKinnon conceptualizes AI agents not as simple tools but as a fundamentally new identity category. This identity possesses attributes of both a human user (roles, permissions) and a system (automation, being headless). This reframing is central to building the next generation of enterprise security and access management.