-1.png)
Ai Governance | Rachel Jannaway | August 2026
A tenant reports damp in their home. A system logs it, scores it low priority, and closes it as a routine repair. No human ever sees the case. Nobody did anything wrong, the system did exactly what it was built to do, ten years ago, for asset management rather than tenant welfare. The thresholds were set by people long since gone, and nobody thought to revisit them when AI was added on top.
What the system didn't know was that the tenant was vulnerable. That fact lived in a free text field, and free text fields don't feed algorithms. A real person, with a real need, slipped through a gap nobody knew existed. This isn't a story about a bad policy. It's a story about an invisible decision, made in the gap between what governance says on paper and what actually happens day to day.
That gap is wider than most boards realise. Recent sector research puts AI use among UK charities at around 79 per cent, most of it informal and ungoverned, while only a handful of trustees believe their charity is using AI at all. The pattern repeats in housing and the public sector: systems quietly running on outdated assumptions, at increasing scale, with almost nobody watching.
The reason this keeps happening isn't a lack of ethics frameworks or policy documents. Most organisations have those. The reason is that governance gets aimed at the document, not the decision. A policy tells you what should happen. It doesn't tell you who is actually seeing a case, who could override it, or whether a complaint gets routed straight back through the system that caused the problem in the first place.
The shift that matters is small but decisive: stop asking "do we have a policy?" and start asking "could I explain this decision to a specific person, doing a specific job, on an ordinary Tuesday morning?" If you can't, you haven't governed it, you've just written about it.
Applying that test consistently changes what becomes visible:
- You know where AI is actually shaping decisions in your organisation, not just where a policy says it should be.
- You know who can genuinely intervene when something goes wrong, and whether they know it too.
- A complaint gets checked by someone outside the system that produced it, rather than looping back through it.
- Your board can explain, in plain terms, why a specific decision affecting a specific person turned out the way it did.
Governance that stops at the policy is governance aimed at intention. Governance that reaches the decision is governance aimed at practice, and it's the second one that actually protects people.
Click 'Read More' to continue reading...