New research from InterSystems puts numbers on a gap that enterprise technology teams have been describing anecdotally for two years: software vendors are confident they've built AI fit for enterprise use, and enterprises keep finding it isn't.
Survey data published today by InterSystems, drawing on responses from 300 independent software vendors across the UK and Ireland, finds 91% of ISVs believe they are prepared to meet enterprise AI demand. At the same time, 97% report difficulties embedding or integrating AI into enterprise environments, and nearly half say their customers are frustrated by the accuracy of AI outputs.
The numbers around governance sharpen the picture. Two-thirds of ISVs plan to increase AI investment over the next twelve months. Only 31% are prioritising governance in their product roadmaps. Yet when asked what enterprise customers value most in AI-powered solutions, 41% name governance — making it the most-cited priority by a significant margin.
Enterprises want AI solutions they can trust – built on compliant, auditable, and high-quality data & platforms that are designed to perform at scale. Reliable, clean data and a scalable architecture are fundamental ingredients to enable AI that performs reliably in enterprise environments
Data infrastructure is where the gap shows up most concretely in the field. More than half of respondents (52%) say their customers' data architectures are actively blocking AI adoption, while 37% identify data quality or availability as a major barrier. Both problems limit scalability, increase time-to-value, and slow the kind of iterative improvement AI products depend on.
In response, 39% of ISVs have integrated a data fabric architecture into their own technology stacks — a higher proportion than might be expected, and a signal that vendors who have done the data infrastructure work are pushing ahead.
"Enterprises want AI solutions they can trust — built on compliant, auditable, and high-quality data and platforms that are designed to perform at scale," said Colin Henderson, Country Manager UK and Ireland at InterSystems. "Reliable, clean data and a scalable architecture are fundamental ingredients to enable AI that performs reliably in enterprise environments."
The research frames the market moment usefully: ISVs racing to add AI features while skipping governance are building product for a procurement conversation that enterprise IT teams have already moved on from. The vendors connecting AI investment to data foundations and auditability are closer to what procurement teams are actually looking for.
