Artificial intelligence (AI) may have become a boardroom priority across global enterprises, but outdated technology infrastructure is emerging as the biggest obstacle to turning AI investments into scalable business outcomes.
A new report by Tata Communications and Bloomberg Media Studios suggests that while enterprises are aggressively pursuing AI strategies, many remain constrained by legacy systems, fragmented technology environments and skills shortages that limit their ability to scale deployments effectively.
The report, titled Building Durable AI Advantage, found that 77% of enterprise leaders now view AI as a board-level priority. However, 65% continue to operate on legacy or developing infrastructure that was not designed to support the data-intensive requirements of modern AI applications.
AI ambition outpaces readiness
The study, based on responses from 501 senior executives across North America, Europe and Asia representing enterprises with annual revenues exceeding $500 million, highlights a growing disconnect between AI ambition and operational readiness.
Only 29% of respondents said their existing infrastructure can scale with evolving business requirements, raising concerns as AI workloads become increasingly complex and resource-intensive.
The report argues that AI success is no longer determined solely by access to algorithms or models. Instead, enterprise performance increasingly depends on the ability to modernise infrastructure, integrate systems, build internal capabilities and establish governance frameworks that support rapid deployment.
Five factors separating AI leaders from laggards
The study identifies five interconnected pillars that determine whether AI investments generate sustained business value: infrastructure, integration, skills, governance and return on investment.
According to the report, weaknesses in any one of these areas can slow enterprise-wide AI adoption and reduce the effectiveness of technology investments.
Infrastructure remains a major concern, with fewer than half of surveyed organisations reporting fully modernised data architecture, network connectivity or hybrid deployment capabilities.
Notably, enterprises with advanced infrastructure were nearly twice as likely to report achieving significant business value from AI compared to those operating on legacy systems.
Integration challenges persist
Connecting AI systems with existing enterprise technology stacks continues to be another major hurdle.
Nearly 28% of respondents cited difficulties integrating AI with legacy systems as a primary barrier to value creation, while 38% said integration concerns contribute to delays in procurement and investment approvals.
The findings suggest that many organisations are discovering that successful AI deployment requires seamless interaction between data systems, applications, workflows and human decision-making processes.
Talent shortage remains a bottleneck
The research also highlights growing concerns around workforce readiness. Around 30% of enterprises identified skill shortages and the lack of specialised AI talent as major barriers to extracting value from AI investments.
The challenge becomes more pronounced among larger organisations. Among companies generating more than $5 billion in annual revenue, 45% cited talent shortages as a significant constraint.
Governance slowing decision-making
As enterprises expand AI deployments, governance processes are becoming increasingly complex.
Security and compliance reviews emerged as the biggest cause of project approval delays, cited by 42% of respondents. Integration concerns and procurement complexity followed closely at 38% each.
The report warns that governance structures designed to reduce risk could inadvertently slow innovation if decision-making processes become overly cumbersome.
ROI remains difficult to measure
Despite widespread investment, many organisations continue to struggle with measuring AI’s business impact.
While nine out of ten enterprises reported some value from technology modernisation initiatives, more than 60% said they have yet to achieve optimal outcomes.
The report suggests that fragmented reporting structures often prevent organisations from capturing the full impact of AI investments across business functions, making it harder to justify additional spending and scale successful initiatives.
Infrastructure becoming the real AI differentiator
Tata Communications believes the next phase of enterprise AI adoption will be shaped less by access to AI technologies and more by the strength of the underlying digital infrastructure.
Tata Communications, President & Chief Revenue Officer, Sumeet Walia, said, “AI has become one of the defining business priorities of our time, but the real differentiator is no longer AI itself — it’s the infrastructure and integration that enable AI to deliver value at scale. Our research shows that while enterprise ambition is accelerating, readiness remains uneven.”
He added that AI is increasingly driving convergence across compute, connectivity, platforms and digital infrastructure, creating a tightly integrated technology ecosystem.
As enterprises move beyond experimentation and pilot projects, the report suggests that the organisations most likely to gain a lasting competitive advantage will be those that invest in the foundational systems required to support AI at scale.