← All events
PAST EVENT / SPEAKER
Building Secure AI Agents
At DevFest, I will discuss how to keep an AI agent within its intended limits when it can access tools, APIs, and data.
MY ROLESpeaker
DATEOct 10, 2026
LOCATIONSanto Domingo
VENUEDevFest · Salón B, Biblioteca Nacional Pedro Henríquez Ureña
About this event
This talk explores trust boundaries in AI agents: how malicious instructions in external content can affect their actions, and which controls reduce that risk.
Topics
- Direct and indirect prompt injection
- Least privilege across models, tools, and data
- Validation, human approval, and action logs