AI risk and resilience
AI risk and resilience: agent incidents, operational resilience, controls, governance and the law.

The OpenAI–Hugging Face incident broke three assumptions in how operational resilience is actually run
Detection worked. Attribution worked. A correct diagnosis on 27 June was followed by a decision not to stop, and a fortnight later Hugging Face was breached.

AI agents escaped containment: 17,600 actions, and the alert that fired too quietly
Three incidents, one headline, three very different severities. What the primary reports actually say, and what to do about it.

I Leaked an API Key and It Turned My Own Domain Into a Phishing Weapon
I got phished from my own verified domain, using an API key I'd leaked myself: what actually leaked, how fast it happened, and what I changed.
I Let My AI Agent Read My Email. Then I Tried to Hack It.
How I built a zero-cost, air-gapped email triage system — and why every AI agent setup with email access is a ticking time bomb.
I Let My AI Agent Read My Email. Then I Tried to Hack It.
Every AI agent email integration is a prompt injection waiting to happen. I built a zero-cost air-gapped system, then attacked it myself. Here's what happened.
I Gave an AI Agent Shell Access to My Server. Here's How I Sleep at Night.
The honest security guide for AI agents. Prompt injection, credential management, model choice, and every mistake I made this week — so you don't have to.