Hi, I'm Irad 👋🏻
I write about AI, security, and software engineering - plus the occasional rabbit hole I swore would only take five minutes.
Recent Posts
Agentic AI in Cybersecurity: Moving Toward Fully Autonomous Pentesting
Hot take: fully autonomous pentesting isn't a moonshot but an engineering discipline. the real question is how can we make HITL the exception?
When AI ‘Snitches’: Whistleblowing Models, Safety, and What Comes Next
Claude 4 and Grok 4 show a new class of ‘high‑agency’ behavior, here's what that means for privacy, alignment, and defensive design.
Proactive Security: Make Systems Secure by Design
Security has been focused on reactive posture management and alert triage. The real question is how do we make vulnerabilities unshippable by design?