Business leaders focused on customer experience (CX) have begun to turn their attention to agentic AI to drive impact, but many still struggle with necessary data access and a lack of confidence in their organizations’ ability to create effective governance frameworks, according to a new survey commissioned by Teradata.
Teradata recently tapped global B2B market research company NewtonX to survey more than 500 AI-relevant executives in companies with at least 100 employees and more than $500 million in annual revenues in industries spanning financial services, government, healthcare (including life sciences), manufacturing (including automotive), retail/wholesale, and telecommunications.
The survey found a majority (74%) of executives have moderate to great confidence in AI’s ability to improve CX, yet they wrestle with the operational realities of effectively implementing it. They generally trust AI’s potential to improve CX, but they still face significant practical challenges with implementation and governance needed to ensure reliability and compliance.
Business leaders now face a paradox in a landscape where CX is a key differentiator. While they and their organizations might be ready to invest in cutting-edge AI for CX, a lack of consistent data and cautious "prove-it-first" mindsets slow agentic AI adoption or hold them back entirely, preventing them from reaping all the rewards agentic AI has to offer.
Organizations favor calculated risk and validation
When it comes to AI adoption for CX, more than half of survey respondents said they are willing to get in early to begin building a competitive position. Yet companies in 2025 also remain measured in embracing new solutions. The data reveal more than a third (35%) now prefer to wait until AI solutions are proven—up from 22% in 2024.
Instead of rushing to be the first-movers, organizations and their executives increasingly want to see innovations validated before committing. Many are more comfortable with the idea of balancing innovation with risk management, rather than simply chasing the AI hype seen in recent years.
Agentic AI has started to achieve significant market awareness and interest. Nearly 85% of organizations polled are exploring or testing agentic AI for CX, while only 9% have fully adopted it. Unsurprisingly, organizations that currently use AI report the highest confidence levels in agentic AI’s ability to improve CX, with 81% very or extremely confident, underscoring the benefit of moving from exploration to implementation.
Yet even as organizations across the world have become more comfortable with AI, executives are still working to balance adoption of advanced AI technologies like agentic AI with CX solutions they can trust.
Data, governance, and skills gaps are key barriers to adoption
What fuels this cautious attitude that prevents some executives and organizations from taking full advantage of agentic AI for CX? The three main issues that surfaced within the survey were data, governance, and a skills gap.
Finding the right data to drive AI for CX is a problem for many. Only 4% of those surveyed reported the right people consistently and quickly get the data they need. On top of that, a whopping 96% experienced some level of delay or inconsistency.
The data issues go beyond supply meeting demand. Accuracy emerged as the top concern (34%) by a narrow margin over security (32%), indicating stakeholders harbor more concerns with the correctness and reliability of data than with its protection. Both are crucial, so organizations must shore up and get a better handle on their data overall for effective adoption.
The research also found governance is a nearly universal obstacle. An overwhelming majority (93%) of executives surveyed faced some level of challenge in creating governance and guardrails for AI initiatives. Organizations with more than $20 billion in annual revenue reported the highest percentage of significant challenge (42%), suggesting governance complexity increases substantially at the largest enterprise scale.
If data and governance weren’t hard enough to overcome, survey respondents further noted an issue with their internal teams and existing skill sets. Only 13% said they currently had the skills needed to advance AI initiatives. In the very near future, organizations must find and hire the right people to help them achieve their CX goals using agentic AI.
Bold AI adopters already see CX benefits
When organizations do find ways to mitigate their data, governance, and skills issues, they’ll have the power to improve their CX function in ways that could even surpass their current goals and expectations.
Virtually all (99%) of the executives surveyed said their organizations have already implemented some form of AI, and nearly 66% expected to see results within six months or less. Their optimism is rooted in perceived reality, as more than half of survey respondents expected to generate at least $1M in either revenue or cost savings from their AI for CX initiatives. That figure includes 36% who expected to create more than $2.5M in benefits.
While internal data, governance and regulatory issues currently loom large over industries, the survey shows business decision-makers overwhelmingly expect agentic AI will improve CX. But for businesses to capitalize on AI for CX, they must overcome those issues. Those that do, however, might find significant returns in revenue, cost savings, and brand reputation.
To learn more about how Teradata enables actionable customer signals with innovative AI solutions, see our Customer Intelligence Framework.