How AI will collide with data readiness

In recent years, the hype around generative AI tools and agentic AI has convinced many leaders to invest big and jump headfirst into the latest advances of the technology without necessarily considering the big picture.

Now that projects are moving from pilot into full production, I expect a lot of these businesses to begin to realize that their data isn’t even close to being AI-ready.

In many cases, the limitations have little to do with the AI itself. Instead, they come from fragmented data, disconnected systems, and foundations that were never designed to support automated decision making or data being shared and acted on in real time.

As AI becomes more integrated into everyday operations, these weaknesses are no longer easy to work around, and they directly impact whether AI delivers value or simply builds cost complexity on top of existing systems.

When AI capabilities outpace data infrastructure

This can be seen in the way AI is being deployed across many organizations, particularly with conversational front ends. They are introduced quickly, often with the aim of reducing friction or improving efficiency.

However, behind the interface, the data being captured doesn’t always flow cleanly into the systems that run the business. In some cases the data is duplicated, and in others it is either incomplete or out of sync with existing records.

This results in AI introducing additional work rather than removing it, with employees spending time checking outputs or correcting errors that originate elsewhere in the system.

While this may have been manageable as a pilot project, as AI moves more into day-to-day operations, these issues become harder to contain — and far more costly.

A clear example of this has been seen in recent AI-driven GP appointment systems. These tools appear effective on the surface, helping patients navigate booking processes more easily, but behind the scenes, the up-to-date patient context and information isn’t always being properly forwarded to the backend GP systems that clinicians rely on.

Not only does this lead to all sorts of data duplication issues and repeat workload for GPs, but it also creates frustrations for the very people the systems have been designed to support.

It’s a classic case of organizations adopting clever AI front-ends without integrating them effectively with backend data and legacy systems, or adopting the operational processes needed to fully realize the value.

Instead of chasing AI features, businesses should start with the outcomes they actually want and work backwards from there. That means focusing on clean, trustworthy data with full lifecycle and lineage visibility, and ensuring it can be acted on in real time.

From big data to fit-for-purpose data

For a long time, data strategy focused on scale. The priority was collecting as much information as possible and storing it cheaply, with the assumption that value could be extracted later.

That approach starts to fall apart once AI is involved because it relies on data that is current and consistent, not hours or days out of date. Outdated or unvalidated legacy records (like old contact details or incomplete customer histories) undermine accuracy and trust in AI outputs.

To get meaningful results, businesses need to prioritize data lineage, governance and context alongside how quickly that data can be accessed and used.

Typically, improving data quality and integration is often seen as a difficult and expensive task, particularly when legacy systems are involved. As a result, many organizations postpone it in favor of more visible AI initiatives.

However, in practice, this delay usually creates more cost over time. Teams spend increasing effort reconciling data, correcting errors and explaining inconsistencies in AI driven outputs.

The opportunity cost is harder to measure but just as significant. When AI cannot be trusted to work reliably, it remains limited to narrow use cases — and without high-quality data foundations, even the most advanced AI initiatives will fall short.

What will change in 2026

In 2026, many organizations will reach a point where improving data quality and integration is no longer optional if AI is expected to deliver meaningful results.

For organizations that want AI to deliver real value, the focus needs to shift away from flashy features and toward fundamentals. That starts with being clear about the outcomes AI is expected to support and working backwards to the data required to achieve them, including how that data is captured, processed and shared in real time.

Data quality, integration and visibility across systems need to be treated as core operational concerns rather than technical clean-up work. Just as importantly, ownership of AI initiatives must be clear.

When responsibility is split or vague, problems in data and process are easier to ignore — getting leadership, IT teams, and frontline staff aligned is essential.

As AI becomes more commonplace across the business world over the next year, those that fail to strengthen their data practices risk ending up with AI that looks impressive on the surface, but delivers little value.

We've featured the best AI website builder.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro