AI security breach 2025, AI guardrail bypass, Aaron Portnoy, Mindgard, healthcare AI security, clinical AI
CybersecurityAI & Technology
DECODE lays out the six-level path from starting out to leading the room, so you can see where each story fits. Free, no email, instant download.
Download DECODEAaron Portnoy's team at Mindgard has tested AI systems across every major industry. Their track record: 100% vulnerability rate. Every system. Every guardrail bypassed.
In this episode of WakeUpX, Aaron makes a bold prediction — a major, high-profile AI application will be publicly breached this year. The industry hasn't had its wake-up call yet. But it's coming.
He also breaks down why healthcare is the sector most vulnerable to real human-level harm from AI security failures. His team has already uncovered vulnerabilities in clinical AI tools, including an application authorized to prescribe medications in the state of Utah.
We cover:
The 100% AI vulnerability rate his team has found across all tested systems
Why no guardrail has held up under testing
His breach prediction for 2025
Why healthcare AI carries a completely different level of risk
Real-world clinical AI vulnerabilities already discovered by Mindgard
Full episodes: www.wakeupxpodcast.com
Mindgard research: mindgard.ai
Aaron's site: aaronportnoy.com
Topics: AI security, healthcare AI, AI breach, guardrail bypass, offensive security, Mindgard, Aaron Portnoy, clinical AI, patient safety, WakeUpX podcast
CISOs, founders, hackers, and people who started last year. The mic is open.