Field knowledge, no buzzwords.
Articles on AI governance, AI security, privacy, AI engineering and IoT asset tracking, from Axon Factory's consultancy and product practice.
Vibe coding: the app works. Can it go live?
With Lovable, Bolt or Replit an employee builds a working app in an afternoon. What a prototype does not show, five questions before go-live and where we help.
Peter Hinssen's The New Normal: what a 2010 book says about AI in 2026
Halfway through the digital revolution, and zero tolerance for failure once something is normal. Why those rules now apply to AI, and what an SME should put in place.
How is your business found in ChatGPT, Copilot and other AI assistants?
What decides whether your business is in an AI assistant's answer: crawlers, readable pages, consistent business details and mentions elsewhere. And how to measure it.
When does your business need a DPIA?
Required for a high privacy risk. The Dutch list, the nine criteria, SME examples and what to do when a customer asks for a DPIA.
PHP2Shell: what this week's PHP flaw means for your website
Two critical PHP flaws let attackers reach web servers without logging in. What to do now, what to ask your hosting provider and what a scan can and cannot see.
Microsoft 365 Copilot is coming to your business: what do you agree beforehand?
Copilot works within existing permissions. First check what is shared too widely, then agree on data, checking output, recordings and incidents.
A larger language model gave worse answers. Measure it on your own documents.
On 39 fixed questions about real fund documents the smallest model scored best. How to measure it, and why one test run is not enough.
Choosing a data room: zero trust document access and monitoring
Who can see what, how you know who saw what, and what if something leaks anyway. With questions for a data room provider.
Glossary: AI governance, AI security and data rooms
AI register, risk classification, shadow AI, RAG, permissions before retrieval and more, each in a few sentences.
AI chat over confidential documents: permissions before retrieval
A RAG chat rarely leaks through the model, but through the search step, the suggestions and the PII masking. What we learned in practice.
Building software with a team of AI agents: what stays with the human
Roles, review, a decision log and tests on what a person sees. And the decisions you don't leave to an agent.
EU AI Act: what is actually expected of you on August 2, 2026
Enforcement starts in days. What must demonstrably be in place: AI register, risk classification and documentation.
Detecting shadow AI without a proxy or endpoint agent
The admin APIs of OpenAI, Anthropic, Microsoft and Google reveal unapproved AI tools, no network project required.
IoT asset tracking without vendor lock-in: why multi-protocol wins
Trackers, SIMs and platform from one vendor is the fastest start and the costliest exit. What to check.