BOSTON — New research finds that internal organizational challenges—not the technology itself—are the primary barrier preventing enterprises from fully realizing the benefits of artificial intelligence.
According to the latest Enterprise AI Benchmark Report from PYMNTS Intelligence, more than seven in 10 executives at large firms said internal constraints are limiting AI performance, while just 11% pointed to shortcomings in the technology.
The findings suggest that while AI is broadly meeting or exceeding expectations, enterprises remain unprepared to deploy it at scale due to issues tied to data, governance and organizational structure, PYMNTS Intelligence reported.
Key Findings
- Internal readiness is the bottleneck. More than 70% of executives said organizational constraints outweigh technological limitations, with only 11% citing AI functionality or accuracy as the primary issue.
- There is no quick fix. Enterprises face multiple, interconnected barriers, including data quality (63%), budget constraints (49%), governance processes (48%) and unclear ownership structures (46%). Executives reported dealing with four to five obstacles simultaneously on average.
- Confidence exceeds reality. While 99% of executives expressed confidence in their data governance frameworks, 85% said their data remains fragmented or only moderately integrated, limiting scalability.
PYMNTS Intelligence said this disconnect highlights a growing “readiness gap,” where companies believe they are prepared for enterprise AI but lack the infrastructure to support widespread deployment.
AI Exceeding Expectations, But…
The report found AI is outperforming expectations across all measured business objectives, with at least 62% of executives reporting better-than-expected performance in every category.
However, PYMNTS Intelligence said that success has shifted the challenge from proving AI’s value to integrating it effectively across organizations.
“This 7-to-1 ratio represents a decisive shift,” the report said, noting the constraint is now the ability of enterprises to adapt their people, processes and data environments.
No Single Solution
Executives surveyed by PYMNTS Intelligence did not identify a single dominant barrier. Integration with existing systems was the most commonly cited top issue, but only by 19% of respondents, followed by data quality (14%), change management (12%), unclear ownership (12%) and budget constraints (12%).
The relatively even distribution underscores that “there is no silver bullet,” the report said, with challenges often interconnected.
Uneven Adoption Across the Enterprise
The research found AI adoption remains uneven across business functions. Advanced AI tools, including large language models and agentic AI, are most deeply embedded in data and technology operations, where 37% of firms report full integration and another 46% report broad deployment.
Other areas lag significantly:
- Growth and revenue functions: 13% deeply embedded, 46% broadly deployed
- Product and customer experience: 11% and 34%, respectively
- Payments and finance: 18% embedding or broadly deploying
- HR, corporate strategy, risk and supply chain: largely in early or pilot stages
PYMNTS Intelligence said these gaps illustrate where organizational readiness challenges are most pronounced.
Closing the Readiness Gap
Despite near-universal confidence in governance frameworks, only 15% of executives described their data as mostly integrated, while 65% said integration remains incomplete and 20% acknowledged significant fragmentation.
To address the gap, PYMNTS Intelligence said enterprises must focus on operational integration rather than policy frameworks alone, emphasizing:
- Stress-testing governance systems against real-world data accessibility
- Prioritizing data integration and system connectivity
- Addressing multiple organizational barriers simultaneously
- Shifting focus from evaluating AI tools to improving organizational readiness
“The gap between confidence and reality is stark,” PYMNTS Intelligence said, adding that AI has effectively outpaced the organizations deploying it.
Methodology
The report is based on a survey conducted by PYMNTS Intelligence of 65 senior technology executives at U.S.-based enterprises with at least $1 billion in annual revenue between Feb. 12 and Feb. 27, 2026. Respondents represented a range of industries, including manufacturing, retail, construction and eCommerce.
