The headline numbers from the UK's latest AI adoption data are striking. 64% of UK organisations now use AI in some form, up from 52% last year. Progress, on the surface. But dig one level deeper and a more uncomfortable picture emerges: only 24% are using AI for advanced organisational change. And on current trajectory, the UK won't achieve full AI adoption until 2102.

That gap, between experimentation and genuine transformation, is not a technology problem. It is a leadership, culture, and capability problem. And nowhere is it more visible than in social housing.

The pilot trap

95% of generative AI pilots fail to reach production. That figure, cited by Professor Alan Brown in a recent session on UK AI strategy, will resonate with anyone who has watched a promising proof of concept quietly disappear after six months. The reasons are familiar: insufficient buy-in, unclear ownership, tools that don't connect to real workflows, and organisations that weren't quite ready for what they were testing.

What tends to be missing is not ambition. It is what Brown calls a disciplined experimental mindset: structured learning loops, contained scope, clear criteria for what success actually looks like. Random trial and error dressed up as innovation is not a strategy. It produces noise, not progress.

The pressure housing leaders are under

The challenge is particularly acute in social housing right now, because the regulatory and policy environment is moving at a pace most organisations are not built to absorb.

Awaab's Law Phase 1 is live. Phase 2 is coming. HHSRS reform is before parliament. The government's response to the Housing Conditions inquiry sets out a new Decent Homes Standard, a 2030 EPC C deadline, building safety requirements, and a ten-year funding programme. The Housing Ombudsman's latest maladministration report shows the same hazards recurring across the sector, driven by fragmented data, siloed teams, and governance that is not catching failures early enough.

Each of these demands is manageable in isolation. Together, they represent a step change in what it means to lead a housing organisation compliantly and responsibly. And the leaders carrying that weight are often doing so without the operational infrastructure to match.

What we see working

At Alix, we work alongside housing leaders who are trying to close exactly this gap. What we have learned is that the organisations making real progress share a few things in common.

They start with a specific problem, not a technology. AI deployed against a vague ambition to "be more efficient" stalls quickly. AI deployed against a concrete operational challenge, whether that is identifying at-risk properties before a complaint is raised, or giving frontline staff faster access to compliance information, builds momentum.

They invest in the human side as much as the technical side. The 70% of UK government bodies that cite AI skills as a barrier are not all facing a shortage of tools. They are facing a shortage of confidence, fluency, and shared language between leadership teams and the technology being introduced. Upskilling is not a one-off training day. It is an ongoing process of building judgment, not just capability.

They treat data as infrastructure, not an afterthought. Many of the failures the Ombudsman documents are, at their root, data failures: information that exists somewhere but is not surfaced to the right person at the right moment. AI can help with that, but only if the underlying data is trustworthy and accessible.

And they resist the urge to let a pilot sit untested indefinitely. The disciplined approach Brown advocates, momentum over perfection, contained experimentation, working with committed stakeholders rather than pushing change onto reluctant ones, is exactly what separates organisations that learn from those that just accumulate cost.

The strategic question for housing

Brown's analysis of the UK's position in the global AI landscape identifies a third path between the US innovation-first approach and the EU's rights-first framework: a distinctly British adaptive path that builds sovereign capability where it matters, leads with balanced standards, and learns from disciplined implementation.

For social housing, the equivalent question is whether the sector can move from being a consumer of AI experimentation to a genuine leader in purposeful AI adoption. The conditions are there. The regulatory pressure is real and growing. The case for operational improvement is compelling. The data, while imperfect, exists.

What is needed now is the leadership will to move beyond pilots, the governance to sustain change, and the honest conversation about where AI can genuinely help and where the problem is something else entirely.

That is the conversation we are committed to having with housing leaders. Not because the technology is the answer to everything, but because used well and implemented carefully, it is one of the most powerful tools available to organisations trying to deliver decent, safe homes at a time when the bar keeps rising.