AI Has Made the Data Quality Problem Impossible to Ignore

For data leaders, AI is just amplifying the oldest challenge in the book…data quality.

Most organisations don’t need convincing that data matters, that argument has mostly been won. The big question is why, after more than 20 years of disruptive data trends, so many businesses are still struggling with the same issue - poor data quality.

That was one of the big topics of discussion at our summer Data Leaders Roundtable, which built on the findings of our new Insight Report ‘Chief Information Overload’ that was launched at the event.

The roundtable discussion explored the pressures facing data leaders as their responsibilities grow, expectations continue to rise and organisations look to data and AI to deliver faster, better outcomes.

While the data leaders around the table were all at different stages of AI maturity, the same message came through consistently. The biggest barrier to successful AI adoption is not a lack of ambition, tools, technical capability or a willingness to adopt the shiny new thing – it is data quality.

Data leaders have seen this all before - business intelligence, big data, cloud platforms, automation and analytics transformation have all promised to help organisations become more data-driven over the decades.

Every new trend brought new language, new tools and new expectations. But so many of the same issues continue to come up: inconsistent definition, unclear ownership, disconnected systems, poor-quality input and teams that do not see data quality as part of their role.

That raises a big question for every business on the AI journey. If data quality has been recognised as a priority issue for decades, why is it so difficult to fix? And how will you fix it now to make AI implementation and data use a success?

The roundtable attendees suggested that the answer is not just technical. Better platforms, governance and processes are important, but none of these things solve the issue on their own.

Data quality has to be considered a people problem. Genuine improvement depends on whether everyone in the business understands the value of good data, accepts responsibility for individual and team roles in the collection of data and everyone treats data quality as part of everyday work rather than a task owned by the data team.

For data leaders, this is now one of the biggest challenges of AI adoption. Their role is not just to manage data or to drive new technology. It is to influence how the business thinks, behaves and makes decisions around data.

The Role has Outgrown the Job Title

One of the other key themes from the discussion was that the data leadership role has grown well beyond what many businesses were originally built to support.

That influence role is becoming even more important as AI raises the ambition. If organisations want AI to deliver value, data leaders need to help the business see data quality as a strategic priority.

AI is Increasing the Pressure

It wasn’t a surprise that AI led the conversation, but the discussion stayed away from the hype. Leaders talked about AI in much more grounded terms - as an accelerator, a pressure point and a mirror for existing problems.

The challenge is that many organisations are still being pulled towards AI before the foundations are in place.

Contributors described pressure from boards, senior stakeholders and teams across the business to “do something with AI”. Sometimes that pressure comes with a clear use case, but mostly it does not. New tools are introduced because they look impressive, because a competitor is using them, or because someone has seen a product that promises to solve business problems overnight.

As one contributor put it “too many AI conversations still begin with the solution rather than the problem we’re actually trying to solve”.

Poor data quality makes that risk hard to ignore. With AI in the mix, poor data is shaping an automated workflow, informing recommendations, influencing a customer interaction and supporting decisions. The more organisations rely on AI, the more important it becomes that the data is proven, trusted, understood and fit for purpose.

The Data Responsibility Rift

The discussion also reinforced one of the big themes in ‘Chief Information Overload’: the Data Responsibility Rift.

Data leaders are becoming more and more accountable for outcomes, but they do not control all the inputs. Data is created, changed, interpreted and used across the business and quality issues sit in day-to-day processes, systems or long-standing workarounds.

The data team are being asked to fix the problem, but the behaviours that created the problem sit somewhere else.

Roundtable contributors were clear that this cannot be solved by technology and unless ownership is understood across the organisation, the same issues will continue to show up.

One of the most practical themes was the need to make data quality meaningful to the people creating and using it. A missing phone number is not just an incomplete field. It means someone cannot be contacted in an emergency. An inaccurate address affects service delivery.

When people understand that connection, data quality can become part of business success.

Culture is the Foundation that Holds Everything Together

Changing behaviour requires patience, clarity and change leadership.

That is where the role of the data leader continues to evolve. Technical capability remains the key, but communication and influence are now just as important. The data leader has to connect the server room to the boardroom, but also the boardroom to the people entering, managing and using data every day.

From Overload to Clarity

Neither the roundtable or the report suggest that businesses should slow down or step back from AI. Far from it, there was clear agreement that businesses cannot afford to ignore it.

But the discussion did challenge the idea that AI success depends mainly on tools.

For AI to deliver meaningful outcomes, organisations need trusted data foundations, clear ownership, practical governance and a shared understanding of what good data actually looks like. Without this, AI risks becoming another layer of noise in an already overloaded environment.

The opportunity for data leaders is to use this moment to change the conversation. AI has made the value of data more visible, but it has also made the weaknesses harder to ignore. That gives data leaders a stronger platform to make the case to build the foundations that every business desperately needs.

What businesses need now is clarity. That clarity starts with data quality, and data quality starts with people.

Interested in the wider challenges facing today's data leaders?
Chief Information Overload explores the growing pressures around data, accountability, AI adoption and decision-making. Read the full report to see what data leaders are experiencing and where organisations need to focus next.