Why organisations are struggling to scale
The research shows that many housing providers have moved beyond initial experimentation with AI. Access to tools is increasing, practical use cases are emerging and organisations are beginning to report benefits. However, organisation-wide adoption remains relatively uncommon and many respondents identified significant barriers preventing wider implementation.
When respondents were asked what is preventing wider AI adoption, skills emerged as the most frequently cited barrier, selected by 35 respondents, 53%. Legacy systems followed closely at 32 respondents, 48%, while budget and data quality constraints were each selected by 26 respondents, 39%. Risk appetite and organisational culture were highlighted by 23 respondents, while 22 identified integration challenges as a significant obstacle.
The findings show that organisations face a combination of workforce, data, technology and organisational barriers rather than a single challenge.
"The findings show that organisations face a combination of workforce, data, technology and organisational barriers rather than a single challenge."
Prevention of wider AI adoption (top 5 selected)
Skills
Legacy systems
Budget
Data quality
Risk appetite
Skills remain the biggest challenge
The prominence of skills as a barrier appears consistently throughout the research. Twenty-four respondents reported that no formal AI support had been provided to staff. Only four organisations reported mandatory training. The most common forms of support were optional training and AI champions or communities of practice, each selected by 21 respondents.
Confidence levels were also mixed. Half of respondents were neutral when asked how confident staff feel using AI effectively. Eighteen respondents, 27%, said colleagues were not very confident and a further two said they were not confident at all. Just 13 respondents, 20%, described staff as fairly confident.
Taken together, these results suggest many organisations are still building the workforce capability needed to support wider adoption. Even where AI tools are available, staff may not yet have the confidence, experience or practical understanding required to use them consistently.
How confident do staff feel using AI effectively?
Neutral
Not very confident
Fairly confident
Not confident at all
Data quality a major obstacle
Data quality emerged as one of the most persistent themes in the survey.
Twenty-six respondents identified data quality as a barrier to wider adoption, while 18 separately selected poor internal data quality. (These were separate options in a multi-select question and may overlap).
Only 14 respondents, 21%, were fairly or very confident in the quality of their organisation’s data for AI, while 21, 32%, reported low or very low confidence.
Confidence in the quality of the organisation’s data for use with AI
Fairly confident
Very confident
Very low confidence
Low confidence
Neutral
Don't know
Improving data quality was selected as a priority by 39 respondents, 59%, making it one of the two most frequently cited priorities for the next 12 months.
Several respondents raised concerns about poor internal data, fragmented information and the risk of generating unreliable outputs from weak underlying datasets. Others highlighted challenges associated with disconnected systems and incomplete information.
For Housemark, the findings reinforce a fundamental principle: the quality and reliability of AI outputs depend heavily on the quality of the data available to the technology.
Legacy systems and integration challenges
Technology infrastructure also remains a significant obstacle.
Legacy systems were selected as preventing wider AI adoption by 32 respondents, 48%, while 22 highlighted integration challenges. The wider technology-readiness question reinforces this: 48 respondents, 73%, said their core business systems offered limited support for AI or no support at all. Responses also referenced difficulties connecting systems, linking data sources and creating a consistent flow of information across the organisation.
These infrastructure barriers may make more advanced use cases involving automation, cross-system data analysis and real-time decision-making harder to implement. Without reliable connections between systems and data sources, many organisations may struggle to scale AI beyond isolated applications.
Governance, culture and organisational readiness
The survey also highlights the importance of organisational factors.
Twenty-three respondents identified risk appetite or organisational culture as barriers to adoption, while 16 identified governance challenges and eight highlighted a lack of executive support. Although these figures are lower than those associated with skills and data quality, they suggest many organisations are still establishing the structures needed to support long-term adoption.
This is reflected elsewhere in the research. Only seven respondents, 11%, reported having a formal AI strategy, while many described adoption as dependent on individual teams, local champions or informal experimentation.
A readiness challenge, not just a technology challenge
The priorities identified by respondents provide further insight into where attention is focused.
Improving data quality and upskilling staff were each selected by 39 respondents, 59%. Twenty-seven respondents prioritised strengthening governance and 18 selected scaling existing use cases. By comparison, only eight respondents identified investment in new tools as a priority.
Within respondents' priorities, investing in new technology ranked well below improving data quality, developing workforce capability and strengthening governance. This suggests many organisations believe the next stage of AI adoption depends less on acquiring additional tools and more on improving the foundations needed to make effective use of the technologies already available.
Housemark’s interpretation is that scaling AI will require organisations to address skills, data, governance, integration and wider organisational capability alongside technology. As adoption matures, these factors are likely to become increasingly important in determining whether organisations are able to move from isolated experimentation to sustainable, organisation-wide impact.