Insights
Writing on technology disputes, digital evidence and AI governance.
AI governance is an organisational control, not an engineering feature
AI governance only works when the whole organisation owns it. How ISO/IEC 42001 and the EU AI Act translate into evidence across the board, legal, data protection, security, procurement, HR and engineering, and what the evidence looks like when governance existed only on paper.
AI-written code and the standard of a competent developer
AI coding assistants have changed how software is written, but not what production software has to be. Why MVP standard is not production standard, what provenance evidence exists for AI-generated code, and what a reasonably competent team does now.
Reading an SDLC from its artefacts, and why it fails without estate controls
Whether a development team followed its own process can usually be shown from Git, CI and ticketing records. But that evidence is only as trustworthy as the IT estate it runs on: unmanaged devices, missing DLP and absent endpoint protection undermine both the software and the record of how it was built.
A year on: my TEDx talk on AI governance is live
A year after my TEDxIllinois Tech talk on AI governance, a look at which predictions came true: product liability for model outputs, sycophancy and memory, ChatGPT ads, and the EU AI Act timetable.