Results Reveal the Sector Needs to Get Clear on AI’s Purpose
The latest On the AI Pulse survey results suggest the social housing sector risks adopting AI without a clear purpose - or a way to measure whether it is making a difference.
Just 13.7% of respondents to our September survey agreed that their organisation was clear about the problems it wanted AI to solve.
The findings are part of our biannual On the AI Pulse research, which includes surveying landlords on an ongoing basis as well as consulting with its expert Delphi panel.
The results suggest a key challenge for housing: while AI use is growing, many organisations are yet to establish the direction, skills and evidence needed to turn that activity into meaningful improvements.
Around two-thirds of respondents to the surveys reported using AI frequently. But using the technology is not the same as knowing where it can add value, or whether it is improving services.
Our report, On the AI Pulse, draws on two pulse surveys carried out in March and September 2026, plus findings from a Delphi panel of experts, all led by Rob Rowlands Research.
It identifies five important challenges for the sector:
- Finding a clear purpose: only 13.7% of September respondents agreed their organisation was clear about the problems AI should solve.
- Proving the value: nine of 13 delphi panel respondents said evidence of meaningful progress was not being collated and analysed. Tools deployed and time saved do not, on their own, demonstrate better services.
- Building the right capability: effective adoption needs skills and collaboration across operational and digital teams.
- Learning what works: organisations need practical experimentation, proportionate governance and meaningful human oversight, with evidence informing whether to continue, change or stop.
- Recognising existing innovation: organisations may be overlooking the people already finding new ways to improve services.
The report calls on housing leaders to start with worthwhile service problems, build the capability to address them, and evaluate benefits from the outset.
It also challenges boards and senior leaders to ask five questions:
- What are we trying to solve?
- What capability do we need?
- Who decides where AI is appropriate?
- How will we know it is working?
- What evidence would make us change direction or stop investing?