The national AI narrative in our Wishlist survey suggests uniform progress, one of AI being implemented to improve productivity and efficiency. The regional reality, as gauged from our latest technology survey completed by over 900 UK wide professionals, tells a different story. 

Across the UK, AI adoption is widespread, but outcomes, experiences and expectations are uneven. Teams in the South of England report significantly higher rates of successful or progressing AI initiatives compared to organisations in Wales, Scotland, the Midlands, and the North.  

At Propel Tech we found this to be fundamentally interesting and it points to the fact that this is not an issue of AI ambition - it's about software maturity, understanding, data and existing infrastructure.

AI success depends on local needs and system maturity

Large-scale national organisations are dominating and leading conversations, which focus on innovation and productivity.  However, our survey revealed that regional organisations are dealing with on-the-ground software issues, including:

• Fragmented legacy systems
• Inconsistent or siloed data
• Limited access to specialist AI expertise
• Highly regulated, process-heavy sectors

A generic AI platform deployed nationally can not account for these variables. What works in a large, well-funded, tech-forward professional services firm will not automatically succeed in a regional manufacturer, logistics operation, or public body. 

When AI fails regionally, it often exposes underlying software, data and systems architectural weaknesses. In practice, feedback suggests that regional AI pilots frequently surface with:

• Poor historic data quality
• Inconsistent process design
• Fragile legacy integrations
• Security and compliance risks

AI does not create these problems, it reveals them when projects fail, and when AI is layered onto unstable systems, performance suffers. That’s why regional AI success is less about access to tools and more about understanding the “as is,” and setting out the data and systems foundations.

Why bespoke software levels the playing field

Closing the UK’s regional AI divide requires AI solutions engineered around legacy systems and local reality. Bespoke software development enables:

• Gradual AI integration with legacy infrastructure
• Secure data architecture tailored to sector requirements
• Workflow design aligned to actual site operations
• AI embedded into systems built for a purpose, not retrofitted

For manufacturers, that might mean AI-assisted scheduling integrated with existing ERP systems. For leisure, it may mean AI enabled booking platforms designed with CRM functionality and payment controls built in. In both cases, AI works because the software architecture and systems infrastructure support it.

Before scaling AI, assess your foundations

Regional organisations often assume they are “behind” in AI adoption. In reality, many simply haven’t assessed whether their existing systems and infrastructure are AI-ready. Before launching another pilot, ask:

• Is our current software secure and scalable?
• Can our data architecture support AI models?
• Are integrations robust enough for automation?

 Start with a free AI software audit 

We’ll assess your code quality, security, and AI readiness and give you a clear roadmap.