House in order: Intranet content governance for the AI era

Generative AI changes the game for organisational content. Traditional search had a simple failure mode: it came up empty, and that was that. AI doesn’t fail quietly. It confidently gives you an answer, even when that answer is wrong or stitched together from mismatched sources, and people tend to believe it.

That risk worsens when policies are inconsistent, documents are outdated, and AI agents aren’t properly coordinated. And most organisations’ AI governance is laser-focused on legal compliance, while content quality barely gets a look-in.

So what does good governance actually look like? This session walks through it: clear ownership of your content, treating what you publish as carefully curated rather than just “stuff that exists,” and paying special attention to policy content, since it tends to be duplicated everywhere and carries the highest risk when it’s wrong.

We’ll also dig into some practical traps to watch for: AI-generated content quietly becoming a “source of truth” nobody signed off on, confusion around licensing and model choice, and accessibility or tagging issues that cause AI to misread your documents entirely. We’ll wrap up with some UX considerations and a handful of quick wins you can act on right away.