AWS re:Invent 2025 Part 1/4 - The Agentic AI Revolution

AWS announces autonomous AI agents that can work for days without human intervention. The DevOps Agent is an always-on incident responder. The Security Agent understands your application architecture. And Kiro is already used by hundreds of thousands of developers.

This is part 1 of our 4-part AWS re:Invent 2025 coverage series.

KEY TOPICS:
• Frontier Agents: DevOps Agent, Security Agent, and Kiro
• DevOps Agent: 24/7 incident response with human-in-the-loop approval
• Security Agent: Context-aware security from design through deployment
• Kiro: GA autonomous developer agent used internally at Amazon
• Bedrock AgentCore: Policy controls, memory, and 13 evaluation frameworks
• Nova Act: 90% reliability on browser automation workflows
• Verification debt: Werner Vogels' concept for AI code generation risks

NEWS SEGMENT:
• Model Context Protocol (MCP) becomes de facto standard for AI-tool integrations
• Oxide Computer Company requires human approval for all AI-generated code

SHOW NOTES:
Full transcript and links: https://platformengineeringplaybook.com/podcasts/00049-aws-reinvent-2025-agentic-ai-revolution