Table of Contents
The study, dubbed ‘Building Durable AI Advantage’, found that 77% of enterprise leaders now consider AI a board-level priority
In sum – what to know:
Infrastructure gap – While 77% of enterprises now treat AI as a board-level priority, 65% still rely on legacy or developing infrastructure.
Scaling challenge – Integration issues, skills shortages, and governance hurdles are preventing many organizations from moving AI initiatives beyond pilot projects.
Readiness divide – Enterprises with advanced infrastructure are nearly twice as likely to report high business value from AI compared with those operating on legacy systems.
While artificial intelligence has become a boardroom priority for enterprises worldwide, many organizations still lack the technological foundations needed to deploy AI at scale, according to a new report from Tata Communications and Bloomberg Media Studios.
The report, dubbed ‘Building Durable AI Advantage’, found that 77% of enterprise leaders now consider AI a board-level priority. However, 65% of respondents said they are still operating on legacy or developing infrastructure that was not designed to support the data intensity and integration requirements of enterprise AI.
The study, which surveyed 501 senior executives across North America, Europe, and Asia at enterprises generating more than $500 million in annual revenue, found that only 29% of organizations believe their infrastructure can scale with evolving business demands.
According to the report, enterprises with advanced infrastructure are nearly twice as likely to report realizing high business value from AI compared with those operating on legacy systems.
The research identified five interconnected areas that determine whether AI investments deliver sustained value: infrastructure modernization, system integration, workforce skills, governance, and return on investment.
Across those areas, the study highlighted several emerging pressure points.
Modernization efforts remain uneven, with fewer than half of enterprises reporting fully modernized network connectivity, hybrid deployment flexibility or data architecture. Integration is also proving challenging, with 28% of respondents citing difficulties integrating AI with legacy systems as a primary barrier to value creation.
Talent shortages continue to constrain AI initiatives. Thirty percent of enterprises cited skills gaps and a shortage of specialized talent as a key obstacle, a figure that rises to 45% among companies with annual revenues exceeding $5 billion.
Governance processes are creating additional friction as well, according to the study. Security and compliance reviews were identified as the largest source of delays by 42% of respondents, followed by integration concerns, and procurement complexity, both cited by 38%.
Tata Communications said the findings point to a growing distinction between AI experimentation and true enterprise readiness.
“The biggest misconception in the market today is that AI adoption and AI readiness are the same thing. They are not,” a Tata Communications spokesperson told RCRTech.
According to the company, enterprises increasingly face an infrastructure challenge rather than an AI adoption challenge. “The AI race is increasingly becoming an infrastructure race. Most enterprises have access to AI models; far fewer have the data foundations, network resilience, security architecture, and integration capabilities needed to operationalise AI at scale,” the spokesperson said.
The company also stressed that organizations seeking to scale AI successfully must move beyond treating AI as a standalone software initiative. “Most enterprises are not building on a greenfield environment. They are trying to integrate AI into complex, existing technology estates,” the spokesperson said.
Network and data architecture modernization will therefore play a critical role in enabling enterprises to transition AI workloads from pilot projects into full-scale production environments, according to Tata Communications.
“Pilot projects often succeed because they operate in controlled, isolated silos. Production at scale, however, demands seamless interoperability and dynamic data mobility,” the spokesperson added.
The report concluded that long-term AI success will depend less on the speed of AI adoption and more on whether enterprises can build the infrastructure, integration capabilities, and operational foundations needed to support AI reliably, securely and at scale.
