Has a lack of data governance become the biggest obstacle to AI?

A study of over 565 data and analytics professionals worldwide reveals a remarkable trend: Data governance is evolving from a neglected, marginal topic to a critical success factor for AI projects. The numbers speak for themselves.

The dramatic increase

In 2023, 27% of surveyed companies still saw data governance as a major obstacle to their data and AI initiatives. By 2024, this number had already risen to 51%. This represents an increase of 89% within just one year.

This development is no coincidence. It reflects the reality many companies are currently facing: the more ambitious the data projects, the more pronounced the governance gaps become.

More programs, more problems

Paradoxically, the number of formal data governance programs is increasing in parallel with the problems. 71% of organizations will have such a program in place by 2024, compared to 60% in the previous year. In the US, the figure is as high as 74%.

This demonstrates a typical pattern: Only when you begin to approach governance systematically does the true complexity become apparent. Defining rules is one thing. Enforcing, monitoring, and scaling them is a completely different matter.

AI as a catalyst

Especially in the context of AI initiatives, governance is becoming a bottleneck. 62% of respondents see it as the main obstacle. There are several reasons for this:

Generative AI and machine learning place higher demands on data provenance and quality. Without traceable data lineage, black-box models emerge that no one trusts.

Regulatory requirements like the EU AI Act are significantly tightening compliance requirements. What was once a "nice to have" is now a legal risk.

Scaling AI applications requires consistent data standards across different systems and teams. Without central governance, data silos and inconsistencies arise.

The concrete benefits are measurable

Companies with effective data governance report tangible improvements:

• 58% see better data quality
• 58% have improved analytics and insights
• 57% experience more collaboration between teams
• 50% achieve better compliance
• 36% have faster data access

These figures show that data governance is not an end in itself, but a direct value driver.

Trust is dwindling

At the same time, distrust in one's own data is growing. 77% of respondents do not fully trust their data for decision-making. In 2023, this figure was 55%. A thought-provoking increase.

The reasons for this are complex: 49% complain about a lack of tools to automate data quality processes. 45% struggle with inconsistent data formats. 43% feel overwhelmed by growing data volumes.

New technologies are gaining importance

The answer for many companies is modern data architectures. Data mesh and data fabric are on the priority list of 18% of respondents, compared to 13% last year. Data catalogs, a new addition to the study, are already a priority for 25%.

These technologies promise greater flexibility and scalability. But they also increase governance complexity. Decentralized data responsibility in the data mesh requires clear ownership models. Data catalogs are only as good as their metadata governance.

The shortage of skilled workers is exacerbating the situation

60% of respondents cite a shortage of skilled workers and resources as major obstacles to AI adoption. The problem: Data governance requires both technical expertise and organizational skills. This combination is rare and expensive.

Many companies underestimate the change management aspect of data governance. It's not enough to implement tools and define processes. People must be brought on board.

What this means in practice

The jump from 27% to 51% in governance barriers is more than just a statistic. It demonstrates that data governance has evolved from a marginal issue to a strategic imperative in 2024.

Companies that take this trend seriously should consider three things:

First, approach governance not only technically but also culturally. The best tools are useless if teams don't use them or circumvent them.

Second, consistently use automation. Policy engines, data lineage tools, and automated quality checks significantly reduce manual effort.

Third, communicate governance as an enabler, not a brake. Teams need to understand that good governance accelerates AI projects, not hinders them.

Conclusion

The study shows a clear trend: Data governance will become critical infrastructure for data-driven companies in 2024. Failure to act now risks not only failed AI projects, but also regulatory problems and a loss of trust.

The good news: The benefits are measurable and the technologies are available. All that's missing is the courage to see governance not as a necessary evil, but as a competitive advantage.

The era of "quick wins" without a solid data foundation is over. In 2024, the companies that have done their homework on data governance will win.

About the study

The study "2025 OUTLOOK: Data Integrity Trends and Insights" was created through a collaboration between Precisely, a global provider of data integrity solutions, and the Center for Applied AI and Business Analytics at the LeBow College of Business at Drexel University.

This partnership is certainly interesting: Drexel University is an established research university (R1 status) in Philadelphia, founded in 1891, known for its practice-oriented education in business and technology. The LeBow College of Business is AACSB-accredited and is considered a renowned business school with a strong focus on applied research.

As a technology company, Precisely brings practical market expertise, while its academic side ensures methodological soundness. This combination of practical relevance and scientific rigor makes the results particularly relevant for practitioners.

However, it's important to keep in mind that, as a provider of data governance solutions, Precisely has a natural interest in highlighting the importance of this topic. The numbers speak for themselves, however, and are consistent with observations from other sources and the daily practices of many companies.

The full study can be accessed here:
https://www.lebow.drexel.edu/sites/default/files/2024-09/drexel-lebow-precisel-data-integrity-trends-insights-2025-outlook.pdf