Past event
The Normalization of Deviance in AI: Lessons From and Beyond the Month of AI Bugs
Virtual event
Feb 19, 2026 6:40 PM
Past event
On-demand webinar
Saving developers from themselves and their AIs
Virtual event
Feb 19, 2025 4:40 PM
Register today
Key Takeaways
Build security boundaries around AI coding toolsAI assistants with access to CLIs, keychains, and code editors introduce risk patterns that traditional AppSec controls were never designed for. See how Lovable defines clear security perimeters that contain AI agent behavior without blocking developer workflows.
Detect and contain rogue AI agent behavior earlyWhen an AI coding assistant acts outside its expected scope, the window to respond is short. Walk through real-world detection patterns that catch unexpected agent actions before they reach production systems.
Ship AI-assisted code faster with practical guardrailsRestricting AI tools slows teams down, but unrestricted access creates blind spots. Learn how to set security defaults that give developers the speed they want while keeping sensitive systems protected.
About this webinar:
AI coding assistants now sit inside developer environments with broad access to credentials, infrastructure, and source code. When these tools act autonomously, they can commit changes, modify configurations, and interact with APIs in ways that bypass existing security controls. This session brings a practitioner perspective on what it takes to protect developers and their AI tools in production. You will see how Lovable's security team built guardrails that contain AI agent behavior, detect anomalies before they propagate, and maintain the velocity that makes AI-assisted development worthwhile. The focus is on patterns you can apply immediately: defining security boundaries for AI tools, monitoring agent actions in real time, and creating response workflows for unexpected behavior
See how to define security perimeters for AI coding assistants that protect credentials and infrastructure without restricting developer productivity
Walk through detection patterns for identifying when AI agents operate outside expected boundaries
Take away a practical framework for building guardrails that scale as your team adopts more AI development tools
For those who are:
Security engineers and AppSec leads who need to protect developer environments where AI coding assistants have broad system access.
Platform engineers and developers building with AI tools who want to understand the security boundaries their teams should implement.
