Care and Lowcode

Before you turn on Copilot, it helps to know what Microsoft’s current Copilot deployment blueprint actually emphasizes. It is organized around three pillars: remediating oversharing, setting up guardrails, and meeting AI‑related regulatory obligations. There is a version of Microsoft 365 Copilot deployment that exists mostly in slide decks. It is clean. Fast. Impactful. Licenses are assigned. People ask questions in Teams. Productivity gains appear neatly in dashboards a few weeks later. And then there is the version that tends to happen in real environments. The one where someone quietly discovers a sensitive document surfaced in a Copilot response because permissions inherited through multiple SharePoint groups were never reviewed properly.  The one where nobody is entirely sure who owns governance decisions once the pilot expands beyond IT. The one where years of “we should probably clean that up later” suddenly becomes operationally relevant. That version is becoming more commo...

I attended a webinar this week that didn’t focus on models or architecture, but on geopolitics. Hosted by the Copyright Clearance Center and moderated by Roy Kaufman, the session featured Anu Bradford , known for the concept of the Brussels Effect . The idea is simple, but powerful: The European Union sets regulation that often becomes a global standard. Why this matters beyond a deadline to How do we build above the minimum? We saw it with GDPR . The argument is we’re seeing it again with the EU AI Act . A lot of discussion around the 2 August 2026 obligations focuses on readiness deadlines. What stayed with me from the session was a different framing: The EU AI Act is not a ceiling. It’s a floor. That shifts the question from How do we become compliant?  To How do we build above the minimum? That feels like a very different motivation - one grounded in trust, not just obligation. Why this matters more than a compliance deadline The 2 August 2026 deadline for high-ri...