Artificial Intelligence (AI) has moved from experimentation to full-scale investment. New data from Semarchy shows that, in the span of a year, the share of C-suite executives allocating more than 20% of their technology budgets to AI has more than tripled – from 16% in 2025 to 50% in 2026.
Today, 97% of leaders report actively investing in AI, up from 74% the year prior. Confidence has risen just as sharply: 92% believe their AI goals are achievable, compared to less than half the previous year, and 98% now describe their organizations as “AI-ready.”
However, early evidence points to an emerging divide. Organizations that run AI on integrated data platforms, implement systematic quality controls, and use Master Data Management (MDM) as a core foundation are beginning to pull ahead. In contrast, those relying on fragmented and poorly governed data are experiencing a very different trajectory marked by project delays, rework, rising costs, and increasing complexity as they attempt to scale their AI initiatives.
The consequences are already visible. More than one in five C-suite executives (22%) report experiencing AI project delays due to data quality issues. A similar amounts cite operational inefficiencies stemming from unreliable data (21%), increased costs from rework or corrections to AI outputs (20%), compliance challenges related to data protection and privacy (19%), and decreased trust in model outputs due to unreliable underlying data (19%).
In short, weak data foundations are eroding the very value AI is meant to create.
Organizations pulling ahead share a common set of practices:
-
Treat data platforms as the engine of AI and agentic initiatives – not a back-office utility
-
Invest in DataOps and operationalized data delivery so data is continuously prepared, monitored, and provisioned for AI
-
Treat data as a product, building reusable, AI-ready data assets
-
Establish MDM as a non-negotiable foundation for both agentic and traditional AI systems
Semarchy’s research highlights five characteristics consistently associated with successful, data-centric AI approaches. The percentages below reflect the share of enterprises that identified each as a key success factor in their AI journey so far:
1. Early governance integration (77%, with 50% in place for over a year)
Leading organizations embed AI considerations into data governance policies and processes early – often well before large-scale deployment.
2. Systematic quality controls (66%)
Organizations actively measure, monitor, and enforce data quality before it is consumed by AI systems.
3. Agentic data management (65%)
Rather than treating AI and data management as separate domains, organizations use data management platforms to directly power AI and agentic initiatives.
4. MDM as a foundation (50%)
Rather than tolerating fragmented data sources, these organizations implement Master Data Management to unify and support AI initiatives.
5. Operationalized data delivery (48% cite as a top investment)
Instead of manually preparing data for each use case, enterprises prioritize DataOps and product-oriented data practices to streamline and scale delivery.
Building the right data foundation is essential for AI success. By investing in these five areas, enterprises can significantly improve their chances of achieving meaningful returns on their AI investments.
