TechAlley AI Meetup
Details
This month we welcom back Matthew Denman for a great talk on practical AI Security. Before we get to the talk, I also wanted to let you know about this upcoming event:
Rebecca Greenhalgh
GTM Strategy & Operations: Airia - Enterprise AI Simplified
[https://luma.com/jksu47fe](https://www.linkedin.com/safety/go?url=https%3A%2F%2Fluma.com%2Fjksu47fe&trk=flagship-messaging-web&messageThreadUrn=urn%3Ali%3AmessagingThread%3A2-Zjc0ZmI5YTItODQ5YS00ZGI1LWFhMjYtNThjZTAyNTJkNzY0XzEwMA%3D%3D&lipi=urn%3Ali%3Apage%3Ad_flagship3_messaging_conversation_detail%3BK6Voo8WOSVedr8p8JdCbiQ%3D%3D)
Matthew Denmans Presentation
Securing AI Agents Isn't a Research Problem - It's Tuesday
Your AI agents can access your payroll system, your customer database, and your infrastructure controls. All from the same conversation. One prompt injection away from chaos. Sound familiar? The good news: we solved this problem in the 90s. It's called RBAC, and most AI teams are ignoring it.
In this session, I'll show you the production-ready architecture that prevents your agents from becoming security nightmares - using bearer tokens, scoped privileges, and tool filtering that works with any LLM framework. You'll see actual code, real implementation patterns, and walk away knowing exactly how to secure your agent systems Monday morning. No theory, no research papers, just battle-tested enterprise security applied correctly. Bring your questions about prompt injection, agent-to-agent communication, and why your 1000-tool context is both a security risk and burning your token budget.