Enterprises are reporting productivity improvements from artificial intelligence, but relatively few are ready to let AI systems make decisions independently across their operations, according to a global study by Tata Consultancy Services (TCS) covering 1,333 senior business and technology leaders in 25 countries.
The TCS Enterprise AI Study: Intelligence that is Real found that 84% of executives surveyed identified productivity improvement as the leading source of value from AI investments. Only 10%, however, said their organizations allowed AI to execute decisions without human approval across multiple functions or business areas.
The findings point to a gap between AI’s growing influence on business processes and the authority companies are prepared to delegate to autonomous systems. While organizations are using AI to support decisions and improve workflows, broad autonomy remains limited by concerns about security, risk, governance and access to enterprise systems.
The study, conducted by the TCS Thought Leadership Institute, surveyed 1,333 senior business and technology leaders across 25 countries and 12 industries. Respondents included C-suite executives, divisional leaders, senior vice presidents and vice presidents from organizations with annual revenues ranging from $500 million to $20 billion or more. enterprise-ai-study-key-findings.pdf
Productivity gains outpace revenue impact
Productivity remains the clearest business benefit of enterprise AI, but the study found that translating those gains into revenue growth is proving more difficult.
Only 10% of respondents ranked revenue growth as the leading source of value from AI investments, compared with 84% who cited productivity improvement. The distinction suggests that companies are seeing more immediate returns through faster work, improved efficiency and employee support than through measurable increases in top-line revenue.
Amit Kapur, TCS chief AI and services transformation officer, said enterprises were moving from using AI to assist employees toward systems capable of taking more autonomous action.
“We are seeing significant progress in scaling AI for enterprises. The shift is clearly in AI moving from assisting people to taking more autonomous action with deep access into an enterprise,” Kapur said.
He added that productivity was where companies were seeing the greatest value, while security, governance and resilience remained major concerns.
“Deploying AI responsibly, with the right guardrails and enterprise controls will be the big focus in the next stage of this journey,” he said.
The study also found that decision-making processes ranked as the area experiencing the greatest AI-related change, ahead of job roles and responsibilities and workforce planning. This indicates that AI’s influence is extending beyond task automation into how organizations evaluate options, allocate resources and make business decisions.
However, influence does not necessarily mean independent execution. The report found that 22% of organizations did not allow AI to execute decisions without human approval. Another 44% permitted it only in a few specific use cases, while 24% allowed autonomous decisions within defined workflows in a single function or business area. Only 10% extended that authority across multiple functions or business areas.
Secure access remains a major obstacle
The limited authority given to AI agents is also reflected in the way companies connect these systems to enterprise applications.
Only 9% of respondents said their AI agents could securely orchestrate workflows across enterprise applications using governed tools and permissions. The largest group, 44%, reported that several enterprise applications and tools were integrated, while 27% had standardized integration platforms enabling secure access across enterprise AI systems.
The report identifies security, risk and governance as the leading considerations shaping decisions about where and how AI systems are developed and deployed. These factors rank ahead of cost, speed and ease of deployment.
The emphasis reflects the growing risks associated with giving AI systems access to corporate applications, sensitive information and operational workflows. As agents move from generating recommendations to carrying out actions, companies must define what systems they can access, what permissions they hold and when human intervention is required.
The report also identified fragmented business knowledge as a barrier to wider AI deployment. Forty-four percent of respondents said important organizational knowledge remained with employees or was dispersed across documentation and systems.
The study’s breakdown showed that 11% said business knowledge resided primarily with employees and was difficult for AI to access, while 33% reported that documentation was fragmented across functions and systems.
Insufficiently integrated, accessible or trusted data was identified as the leading challenge to scaling AI beyond initial use cases. Other obstacles included difficulties getting business teams to adopt and operationalize AI, rising costs, limitations in existing systems and the complexity of managing AI models and tools.
Model independence and sovereignty concerns grow
The survey found that enterprises are also seeking greater flexibility in their AI infrastructure.
Nine in 10 respondents said independence from a single AI model or platform was important to their investment decisions. For 27%, it was a strategic requirement guiding architecture and investment, while another 39% considered it important across most AI initiatives.
AI sovereignty is another growing concern. The study defined this in terms of organizational control over AI data location, model choice and regulatory or data-residency compliance.
Thirty-four percent of respondents classified sovereignty-related requirements as a high or very high organizational risk. The proportion rose to 43% among respondents in continental Europe.
Despite these concerns, organizations remain poorly prepared to move critical AI workloads between providers. Only 7% said they had business continuity plans that would allow them to switch critical AI workloads across their entire organization. Another 34% were prepared to switch providers for most business-critical use cases, while 47% had plans covering only a few critical use cases.
Among organizations with continuity plans, 35% said they had not tested them.
Open-source AI adoption is also selective. Forty-six percent reported deploying open-source models for selected use cases, while 13% said they had deployed them broadly across multiple functions. Seven percent described open-source AI as a strategic priority involving significant customization or deployment.
The study further found that 39% of organizations assess whether AI initiatives deliver sufficient value relative to token and computing costs. Another 31% track these expenses but do not consistently evaluate whether the resulting business value justifies the consumption.
Workforce impact remains uncertain
The report found that executives have become less certain about AI’s long-term effect on employment.
The proportion of respondents who said it was too early to determine how AI would affect job roles increased from 4% in TCS’ 2024 AI for Business Study to 19% in the latest survey.
At the same time, the share expecting AI to eliminate more jobs than it creates declined from 47% in 2024 to 37% in the current study. Those expecting more new roles than eliminated positions also fell, from 32% to 26%.
TCS said the results should not be interpreted as a straightforward shift toward either optimism or pessimism. Rather, practical experience with AI appears to have made its workforce implications harder to predict.
Companies are already changing job responsibilities and workflows, but their ability to prepare employees for those changes remains uneven. Only 28% of respondents said they maintained an enterprise-wide skills framework or an evolving skills taxonomy connecting capabilities with roles, career paths, learning and workforce planning.
The study also found that only 9% treated AI as a strategic enterprise capability supported by standardized architecture, governance, reusable assets and measurable business value. Thirty-eight percent relied on an AI Center of Excellence, while 29% reported some form of shared governance.
TCS said enterprises seeking to scale AI should strengthen data and security foundations, design systems for flexibility and provider independence, establish clear executive accountability and connect workforce planning with changing ways of working.
The report’s central finding is that enterprise AI maturity depends on more than access to models and tools. Organizations must also connect technology with business processes, governance, decision-making and workforce capabilities before they can safely expand AI’s authority.
Full disclosure: All news articles published on the TechSabado website are written by human journalists, unless otherwise specified. Final text editing is also performed by human editors, with artificial intelligence (AI) used only to assist with additional grammar and style guide corrections..
————————————————————————-
TEN YEARS OF TECHSABADO!
PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE & SPOTIFY channels.
WATCH TECHSABADO ON OUR YOUTUBE CHANNEL:
WE ARE ALSO ON SPOTIFY
WATCH OUR OTHER YOUTUBE CHANNELS:
PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE CHANNEL.
Leave a Reply