Recommendations and next steps
The findings from this research suggest that social housing has reached an important stage in its AI journey. AI activity is widespread among respondents, practical use cases are emerging and many organisations are reporting benefits. However, organisation-wide adoption and consistent measurement remain relatively uncommon.
For much of the respondent base, the focus is shifting from whether AI should be used towards how experimentation can be translated into sustainable organisational impact.
Based on the findings, Housemark believes there are five priority areas organisations should consider over the next 12 months.
1. Improve data quality and accessibility
Data quality was one of the most consistent themes throughout the research. Twenty-six respondents identified data quality as a barrier to adoption, while 18 separately highlighted poor internal data quality. Improving data quality was selected by 39 respondents, 59%, making it one of the most frequently cited priorities.
Housemark recommends that organisations assess whether their data is sufficiently accurate, accessible and trusted before attempting to scale more advanced AI applications. Many of the use cases emerging across the sector, particularly those involving automation, prediction and decision support, rely on strong underlying data foundations.
2. Invest in workforce capability
Skills were identified as the most significant barrier to wider adoption, selected by 35 respondents, 53%. Upskilling staff was also selected by 39 respondents, 59%, making it one of the most commonly cited priorities.
The findings show that workforce capability is a major part of the adoption challenge alongside technology, data and organisational factors. Organisations should consider how they can improve confidence, provide practical training and help colleagues understand where AI can add value within their day-to-day roles. Leadership also matters: senior leaders can set priorities, establish clear ownership, create permission for responsible experimentation and maintain accountability for the outcomes AI activity is expected to deliver.
3. Prioritise use cases that solve real problems
Respondents reported a wide range of applications including reporting, data analysis, coding, complaints handling, contact centre support, workflow automation and predictive modelling.
Housemark recommends focusing initially on a manageable number of use cases aligned with clear business priorities or service objectives. Many of the practical examples identified in the research were linked to identifiable operational challenges rather than AI being deployed for its own sake.
4. Strengthen governance and oversight
The research highlights continued concerns around data protection, security, hallucinations, bias and compliance. Twenty-seven respondents selected strengthening governance as a priority for the next 12 months.
Housemark’s view is that proportionate governance can enable adoption by providing colleagues with clear guidance, safeguards and confidence. Governance should support responsible experimentation, not prevent it.
5. Measure outcomes, not activity
One of the most significant findings is the gap between perceived value and proven impact.
While 62% of respondents agreed or strongly agreed that AI is delivering measurable value and 65% agreed or strongly agreed that it has improved productivity, only 21% said they could demonstrate return on investment. Meanwhile, 70% are not measuring AI impact at all.
As adoption matures, organisations should consider how they will evaluate success, identify high-value use cases and make informed decisions about future investment. Without measurement, it becomes difficult to distinguish genuine impact from anecdotal success.
Practical checklist
Start:
- Identifying high-value use cases
- Improving data quality
- Upskilling colleagues
- Establishing simple measurement approaches
Stop:
- Treating AI solely as a technology project
- Assuming access alone will drive adoption
- Pursuing use cases without clear objectives
Continue:
- Supporting controlled experimentation
- Sharing good practice
- Strengthening governance
- Building organisational confidence
The findings suggest that the next phase of AI adoption will depend less on purchasing additional technology and more on organisational readiness. Housemark believes organisations will be better placed to realise long-term value when practical use cases are supported by reliable data, workforce capability, proportionate governance and effective measurement.