A survey of 818 data intelligence, analytics, and infrastructure decision makers conducted by the Futurum Group finds there has been a shift in the objectives teams are prioritizing that focuses more on driving new business opportunities, meeting financial goals and completing projects.

At the same time, there appears to have been a decline in building artificial intelligence (AI) capabilities and efforts to ensure data is trusted. Building AI capabilities (24%) and increasing trust in data (18%) are still the top objectives, but each declined 6% and 7%, respectively, in the past six months.

In comparison, there has been a 5% increase in the number of survey respondents (17%) who cited driving new business opportunities as their most important business objective, followed by meeting financial goals (15%) and completing projects (13%).

The survey suggests the pendulum is now swinging from aspirational architecture to pragmatic delivery, says Brad Simmins, vice president and practice lead for data intelligence, analytics, and infrastructure at the Futurum Group. For example, trust is no longer being funded as its own independent project. Instead, governance is being forced to embed directly into active delivery pipelines, he added.

Overall, business unit leaders, rather than centralized IT, are steering the ship now, noted Simmins. They care about velocity, uptime, and monetizing data assets right now as data teams morph into profit centers, he noted. “Their metric for success is ‘what did we ship this quarter?’ instead of, ‘gosh, look how elegant our data model is,’” says Simmins.

In general, AI is still at the center of most new business and technology initiatives, but it’s also clear that organizations are finding it challenging to incorporate AI into existing workflows. Much of that challenge stems from a lack of data governance frameworks needed to incorporate probabilistic AI technologies into what are often deterministic business workflows that require tasks to be completed the same way each time. Many of those workflows may also need to be re-engineered altogether because they were initially designed for humans rather than AI agents.

Finally, organizations are increasingly experimenting with agentic AI workflows that will be used to realize new business opportunities, while at the same time trying to ensure they remain competitive as AI increasingly becomes a new set of table stakes required to remain relevant.

Regardless of motivation, pressure to realize a return on AI investment is running high as the cost of implementing AI at scale becomes more apparent. While there may not be as much anxiety about the fear of missing out should rivals incorporate AI into workflows faster, there is increased focus on identifying a smaller number of AI initiatives that will deliver business value sooner rather than later. As a result, there is now more focus on both improving the quality of the data needed to drive those initiatives and the controls needed to ensure that AI agents don’t exceed the scope of the mission assigned. After all, it may only take one AI agent to create an incident that negates any of the business value that might otherwise have been derived.

At this juncture, it’s not so much a question of whether to use AI, but rather which initiatives enable an organization to simply remain competitive versus the ones that provide a true competitive advantage that can be quickly realized.