[un]prompted II: Full Coverage

AI security is moving beyond the hype, and [un]prompted II is bringing together the researchers, hackers and security leaders determined to find out what actually works.

Taking place October 27–29, 2026, at The Hibernia in San Francisco, the AI Practitioner Conference features Bruce Schneier, Gadi Evron, Taesoo Kim, Elie Bursztein, Chenxi Wang, Rich Mogull, Kayla Williams, Olivia Rose and dozens of other experts tackling the security challenges of agentic AI, autonomous attacks, vulnerability discovery, governance and production defenses.

The agenda also features three head-to-head debates — Sergej Epp vs. Jeremiah Grossman, Tim Brown vs. Gadi Evron, and Katie Moussouris vs. Stefano Zanero — along with hands-on demonstrations and research presentations.

CYBR.SEC.Media is covering [un]prompted II from start to finish. This page brings together our latest news, speaker insights, research analysis, interviews and conference takeaways, with continuing updates throughout the event.

The story so far:

Who Is Prompt-Injecting Coding Agents? A Scan of 86 Million GitHub Items

At [un]prompted II, Radware researchers reveal how hidden prompt injections found in a scan of 86 million GitHub items can manipulate AI coding agents, redirect workflows and threaten sensitive data.

CYBR.SEC.MediaBill Brenner

From DARPA AIxCC to Microsoft FORGE Lab: Lessons Building Hyperscale AI Code Scanning Tools

Microsoft FORGE Lab's AI agents helped find 140 Windows CVEs. At [un]prompted II, Taesoo Kim explains why fixing bugs is the bigger challenge.

CYBR.SEC.MediaBill Brenner

AI Agent Sandboxing: Google Researcher Unveils Capsem

Google DeepMind's Elie Bursztein presents Capsem at [un]prompted II, using hardware isolation to secure AI agents without sacrificing speed.

CYBR.SEC.MediaBill Brenner

AI Agent Security Gaps Exposed at [un]prompted II

A review of submissions during the CFP found 71% of 490 conference submissions focused on AI agents, exposing gaps in production defenses, vulnerability remediation and threat intelligence.

CYBR.SEC.MediaBill Brenner