CyberWire Daily Dial M for malware.
Oct 30, 2025
Mike Anderson, the Chief Digital and Information Officer at Netskope, joins to redefine how CIOs should approach Agentic AI by thinking like HR leaders. He emphasizes the need for HR-style access controls and guardrails to manage AI agents effectively. The conversation explores the risks of automating flawed processes and the importance of redesigning workflows before adopting AI. Anderson suggests that organizations should start with small, proven use cases to avoid disruption. This insightful dialogue bridges the gap between technology and human resource management.
AI Snips
Chapters
Books
Transcript
Episode notes
Agentic AI Is A Workflow Assistant
- Agentic AI acts as a natural-language front end to workflows, behaving like an assistant for specific tasks.
- True autonomous agents that act without direction are likely still years away.
Treat Agents Like Employees With Limited Access
- Restrict agent access just as you would a new employee and avoid granting super-admin rights by default.
- Apply role-based guardrails and limit sensitive system access for AI agents.
Start With Narrow, Role-Based Agents
- Start agents with narrow roles (travel planner, calendar manager) and define distinct guardrails per role.
- Gradually increase autonomy after agents reliably follow policies and preferences.
