This is where I write about the problems I’m working on. I file everything under three headings: AI, Cybersecurity, and Online Safety & Privacy.
Every other person on that video call was AI-generated, and the finance employee had no way to tell. Training people to spot fakes won’t fix this. The way payments get approved has to change.
The demo works on a clean PDF. Real document sets have scanned annexes, several languages, and tables that don’t mean what they seem to. This is what a system that copes with them looks like.
AI agents now get credentials, tool access and the freedom to act on their own. We spent twenty years learning how to manage accounts like that, and most companies aren’t using any of it.
Process recipes and design files are some of the most targeted IP there is. How to give engineers useful AI tools without your designs ending up in someone else’s training data.
Articles I’ve written and stories that quote me, newest first.
If you’re commissioning a piece or putting a panel together, these are the areas I cover.
AI governance, AI agents as identities, AI-driven attacks, and why AI that works in a demo breaks on real documents.
Security programs that hold up, deepfakes in the approval chain, keeping design IP out of public models, and what I’ve learned as a CISO.
Children online, social media rules for minors, and how ordinary people protect their personal information.
Get in touch if you’d like an article, a comment on deadline, or a speaker for a panel.