AI Security Is Back in the 90s | Aaron Portnoy on Hacking AI, Zero Days & Breaking Into Cybersecurity
CybersecurityAI & TechnologyCareerBusiness & LeadershipEmpowerment
What if the AI powering your company is about as hackable as a 1999 web server?
That's the uncomfortable reality Aaron Portnoy — offensive security legend, co-founder of Exodus Intelligence, and Head of Research & Innovation at Mindgard — lays out in this episode of WakeUpX.
We cover:
- Why AI security mirrors the early internet era
- The "hamster wheel" of repeating security trends
- The state of AI guardrails (and why they're not enough yet)
- How to break into cybersecurity with no degree or certifications
- Why curiosity is the most dangerous skill in offensive security
Aaron has spent 20+ years at the frontier of vulnerability research — from running the Zero Day Initiative to architecting the Pwn2Own competition, building Randori's R&D (acquired by IBM), and serving as the inaugural Hacker Fellow at Dartmouth College. He's now applying that offensive lens to AI systems at Mindgard.
Full episodes: www.wakeupxpodcast.com
Connect with Aaron: aaronportnoy.com | mindgard.ai/authors/aaron-portnoy
Topics: AI security, cybersecurity career, offensive security, zero day research, AI red team, hacking, prompt injection, Pwn2Own, Mindgard, WakeUpX
The free library, certification roadmap, and home lab guide.
See the resources › FIND YOUR DIRECTIONTell us what you are working on and get pointed at the right next step.
Find my path › MORE CONVERSATIONSBrowse or search the full archive.
Browse episodes › CONNECTFind mentors, peers, and the connections you were missing.
Join the community ›CISOs, founders, hackers, and people who started last year. The mic is open.