Artificial intelligence (AI) has become a boardroom priority for enterprises worldwide, but many organisations remain constrained by outdated infrastructure that is ill-equipped to support AI at scale, according to a new global study by Tata Communications and Bloomberg Media Studios. 

The report, titled Building Durable AI Advantage, found that 77% of enterprise leaders now view AI as a board-level priority. However, 65% of organisations continue to operate on legacy or developing infrastructure that was not designed to handle the data intensity, computing demands and integration requirements of modern AI applications. 

Only 29% of respondents said their infrastructure is capable of scaling in line with evolving business needs, highlighting a significant readiness gap as AI workloads become increasingly complex and resource-intensive. 

The study surveyed 501 senior executives across North America, Europe, and Asia from companies with annual revenues exceeding $500 million. 

According to the report, long-term AI success depends on the strength of five interconnected pillars — Foundation, Integration, Skills, Governance and Return on Investment (ROI). Weaknesses in any one of these areas can limit the ability of organisations to extract sustainable value from AI investments. 

Infrastructure modernisation remains uneven 

The report found that fewer than half of enterprises have fully modernised critical areas such as network connectivity, hybrid deployment capabilities and data architecture. Companies with advanced infrastructure are nearly twice as likely to report significant business value from AI compared with those operating on legacy systems. 

Integration challenges also remain a major obstacle. Around 28% of executives cited difficulties integrating AI with existing legacy systems as a key barrier to value creation, while 38% said integration concerns contribute to delays in project approvals and procurement decisions. 

Nearly two-thirds (67%) of respondents said seamless integration between digital automation and human interactions across channels is critical for successful AI deployment. 

Skills and governance challenges persist 

Talent shortages continue to hamper AI adoption, with 30% of organisations identifying skill gaps and a lack of specialised expertise as a major constraint. The challenge is more pronounced among larger enterprises, with 45% of companies generating more than $5 billion in revenue citing talent shortages as a significant concern. 

Governance-related bottlenecks are also slowing AI implementation. Security and compliance reviews were identified by 42% of respondents as the biggest source of approval delays, followed by integration issues and procurement complexities, each cited by 38%. 

The report warned that as organisations create larger stakeholder committees to oversee AI investments, governance frameworks risk becoming obstacles to scaling rather than enablers of innovation. 

Measuring AI value remains a challenge 

While nine out of ten enterprises reported deriving some benefits from modernisation initiatives, more than 60% said they have yet to achieve optimal outcomes. The study noted that many organisations continue to assess AI, infrastructure and security initiatives separately, making it difficult to capture the broader business impact of AI investments and limiting the ability to make informed reinvestment decisions. 

Sumeet Walia, President and Chief Revenue Officer at Tata Communications, said that infrastructure and integration capabilities are becoming the key differentiators in the AI era. 

"AI has become one of the defining business priorities of our time, but the real differentiator is no longer AI itself — it is the infrastructure and integration that enable AI to deliver value at scale," Walia said. He added that organisations best positioned for future growth will be those investing in the foundational systems that connect people, data, platforms and intelligence across the enterprise.