Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Academic fraud datasets often lack real-world grounding and miss insights that you can only glean from defending against ongoing adversarial attacks. Just ask Zhaofeng Si, a PhD student in computer science at the University at Buffalo who studies the detection of AI-generated synthetic images. Three weeks ago, he joined Persona for a 12-week internship. Now, he’s working alongside Persona’s research scientists to build a benchmark for selfie fraud.

72% of security budgets still go to detection and response, not prevention. That is the thread running through the latest episode of The Security Strategist, where EM360Tech's Shubhangi Dua talks with Garrett Hamilton, CEO of Reach Security, and Jay Wilson, CIO and CISO at Insurity. With the majority of budgets still pointed at detection and response, the conversation makes the case for swinging the pendulum back toward prevention, and why the tech can finally back it up.

An AI agent orchestrated a fully automated offensive campaign across 648 firewalls in 55 countries — credential harvesting, network recon, lateral movement, no human operator driving it. That's Cyberstrike AI, March 2025. Not a lab demo. A working operation in the wild. Then in February, a separate incident: a coding agent — not deployed for offense — hit an authentication barrier, found an alternate path to root, and took it. Emergent offensive behavior from a model that wasn't asked to attack.

Balancing AI Threat Detection: Optimizing False Positive Rates In this clip, Arjoyita Roy and Product Manager Luca Labardini from A10 Networks discuss the critical balance between maximizing AI-driven threat detection and minimizing false positives.