Welcome to Tech + Policy

Why a separate section

The Insights section is where I publish long-form thinking on AI strategy, the economics of innovation, and the operating models that make frontier technology stick. That is one register: pragmatic, decision-orientated, written for people who have to ship.

Policy is a different register. The questions are slower. They cross jurisdictions. They sit between what the technology can do, what institutions are built to absorb, and what the law has yet to catch up with. I have been writing about this in fragments across the Insights section and in the work that does not get published. Pulling it into its own space makes the analysis more comparative and the conclusions more honest.

What lives here

Three kinds of pieces, roughly:

1. Regime breakdowns. A single use case (say, frontier-model deployment in financial services, or biometrics in border control) traced through the frameworks that bind it: EU AI Act, NIST AI RMF, the African Union Continental AI Strategy, POPIA and the SA Draft AI Policy, China’s AI Measures, sectoral regulators where they matter. The point is not advocacy. The point is to make the deltas legible.

2. Implementation post-mortems. When a policy lands in practice, what actually happened? The gap between what regulators wrote and what compliance teams shipped is where most of the interesting questions live.

3. Field notes from the work. Thoughts that surface while I am drafting, advising, or reading public consultations. Shorter pieces. Less polished.

The Explorer

At the bottom of this page is PolicyPrism: a working tool that refracts a single AI use case through five governance regimes and shows where the obligations diverge. It started life as a demo. It is now the analytical backbone of this section. If a piece references a comparison, you can recreate it there.

If something here is wrong, ambiguous, or missing the regime that matters most for your work, say so. The point is to think in public, which means being correctable in public.