The Application Security Podcast Francesco Cipollone - Agentic AI Manifesto
5 snips
Sep 23, 2025 Francesco Cipollone, CEO of Phoenix Security and a seasoned expert in AI and application security, discusses the critical distinction between true AI agents and chatbots. He emphasizes the importance of using multiple LLMs for effective outcomes and outlines the key principles of his Agentic AI Manifesto, advocating for human augmentation over replacement. Francesco also highlights the economic challenges of AI implementation and warns businesses against half-hearted approaches, making a case for responsible AI adoption to stay competitive.
AI Snips
Chapters
Transcript
Episode notes
Walk-And-Talks And Mini Whiskey Events
- Francesco started 'cyber walk and cyber talks' by recording short walk-and-talks during breaks with his remote team.
- He runs regularly and collects whiskey minis used for company events.
Agents Should Augment, Not Replace Humans
- Francesco frames the Agentic AI Manifesto to promote agents that augment humans, not replace them.
- He urges applying AI only where logical benefit exists and to keep humans central to decision-making.
What Makes An AI Agent Autonomous
- Francesco defines an agent as autonomous, reactive, memory-equipped, and capable of chaining with other agents.
- He contrasts true agents with glorified chatbots that merely wrap an LLM with a prompt.