Data quality gaps threaten AI progress & compliance in business
New research from Ataccama indicates that enterprises are facing significant challenges in achieving data quality sufficient to support both artificial intelligence initiatives and regulatory compliance requirements.
The Ataccama Data Trust Assessment, which has been completed more than 150 times by chief data officers, business leaders, and professionals in data management since March, provides a benchmark for data trustworthiness within organisations. Results show that organisations average a score of only 42 out of 100 for data trust maturity. The lowest scores are seen in remediation workflows, policy enforcement, and reference or master data quality.
These findings suggest that the accelerated adoption of AI and the rise in regulatory requirements are making weaknesses in enterprise data more visible, weaknesses that slower-moving business environments may have concealed in the past. When data cannot be relied upon, both AI and compliance efforts are at risk of failure.
The assessment was developed to guide organisations from performing inconsistent and isolated data management projects towards achieving a maturity-based, comprehensive programme. This approach addresses people, process, and technology, assigning a baseline score, providing peer benchmarks, and outlining a set of ranked priorities across four pillars: quality, governance, observability, and improvement through remediation and workflow enhancement. Participants are encouraged to repeat the assessment to monitor the progression and strategic scaling of data maturity programmes.
AI and compliance struggles
Ataccama's broader research also finds that while roughly a third of organisations report making meaningful progress with AI, most identify data quality as the primary obstacle. Significant time is spent by leaders in searching for, validating, and reconciling data instead of deploying AI models. The report positions data quality as central to establishing trust - described as the fundamental requirement for AI and compliance to deliver positive outcomes.
There are differences in how these gaps manifest across sectors, but the underlying issue remains consistent: results are compromised without trusted data. In financial services, poor observability and weak data lineage make auditing and tracing back reports to original sources difficult, escalating regulatory risks. For manufacturers, inconsistent product data has a cascading impact, delaying reporting, disrupting supply chains, and increasing costs of compliance. Across all industries, the report concludes that high data quality is essential for generating dependable results.
Leadership perspective
"Quantitative intuition, the mix of data and judgment, drives better decisions. Untrusted data erodes every decision it touches," said Jay Limburn, Chief Product Officer at Ataccama. "Too many organizations invest in data programs without a clear view of how trustworthy their data landscape really is. We built the Ataccama Data Trust Assessment to surface where trust breaks down and to guide what to fix first. Data trust creates value on every front. Without real insight into data quality, businesses risk cascading failures – from unreliable AI outputs to stalled growth. Trust has to permeate every layer, from data and models to decisions."
The Ataccama Data Trust Assessment aims to supply organisations with actionable insights to prioritise and remediate issues, supporting businesses as they work towards robust, scalable data governance. Leaders are able to track their progress over time and direct investments to those areas that have the greatest impact on their data ecosystem.
The research and assessment tool reflect a growing industry focus on establishing strong data quality as foundational not just for technical outcomes like better AI performance, but also for addressing escalating regulatory and compliance obligations that demand auditable and traceable data management practices.
