2026-07-21
The fastest-growing shadow AI isn't your employees using ChatGPT. It's the AI your trusted vendors switched on by default, with no one notified. Under the EU AI Act you own that decision, and you're usually the last to know. Here's how to see it.
2026-07-14
Europe doesn't need its own ChatGPT. It needs muscle in six specific places. Here is the map, what to build versus buy at each layer, and why the honest scorecard matters more than the ambition speech.
2026-07-06
Most AI-and-jobs debates argue at the wrong altitude. Here is why disruption moves through three distinct layers, tasks, roles, companies, and why the real risk sits at the layer almost nobody is watching.
2026-06-30
When output becomes cheap, value moves to judgment. Here are the five skills that compound in an AI-native world, why the fastest wins are boring, and how to spend the AI dividend on time instead of more shallow work.
2026-06-23
The biggest AI risk is rarely on the risk register: the opportunity cost of not using AI well. Here is how to govern outcomes instead of tools, why centralized control fails, and how the EU AI Act is guardrails rather than a brake.
2026-06-16
An Amsterdam data center does not make your AI sovereign. Here is the four-layer map of where control actually lives, the jurisdiction question most vendors skip, and what you should own versus rent.
2026-06-09
Most AI policies are written to pass an audit, not to govern. Here is why bans backfire, what the data says, and the discover-before-you-govern approach that actually reduces risk.
2026-03-04
Why we are building Prometheon, and what you can expect from our new blog dedicated to secure, on-premise enterprise AI and related topics.