Details

The AI hype cycle is over. The Security Hangover has arrived. So who actually has the keys to your AI? Kate Catlin from GitHub and Dan Moore from FusionAuth sit down for a fast, conversational deep dive on the Authorization Paradox — why probabilistic AI agents have to be governed by a deterministic identity layer, what Copilot's security tooling is doing about it in real codebases, and where this is all headed. Stick around for live community AMA.

What You'll Learn:

- Why identity is "Layer Zero" — the infrastructure primitive that must exist before AI agents "wake up"

- How probabilistic AI must be governed by deterministic AuthZ — and what happens when it isn't - A working tour of GitHub Copilot's security tooling: Vulnerability Prevention, Autofix, and custom guardrails via .github/copilot-instructions.md

- How FusionAuth's isolated deployment model contains the "Blast Radius" when an agent goes off-script - Where authorization is heading: from "Is this user logged in?" to "Does this agent have permission for this API call, right now?"