Executives around the world, including in Singapore, are sold on what artificial intelligence can do for productivity. They are far more cautious about letting it act on its own. A new global study by Tata Consultancy Services (TCS) found that 84% of executives rank productivity improvement as the leading source of value from their AI investments, yet just 10% of enterprises let AI agents execute decisions without human approval across multiple business functions — a gap that sums up where enterprise AI adoption stands today: strong on output, cautious on control.

The study, TCS Enterprise AI Study: Intelligence that is Real, surveyed 1,333 senior business and technology leaders across 25 countries and 12 industries, including Singapore. Respondents included C-suite executives, divisional leaders, SVPs and VPs at organisations with annual revenues ranging from US$500 million to more than US$20 billion.

Productivity leads, autonomy lags in enterprise AI adoption

The headline figure in the study is the productivity payoff: 84% of executives point to it as the biggest source of value from their AI spending. Revenue growth is a smaller story so far, with only 10% of executives crediting AI for an uplift.

What enterprises are not yet doing is letting AI run on its own. Only 10% allow AI agents to execute decisions without human sign-off across more than one function or business area, and just 9% say their AI agents can securely orchestrate workflows across different enterprise applications.

“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,” said Amit Kapur, Chief AI and Services Transformation Officer at TCS. “Productivity is clearly where enterprises are seeing the greatest value from AI today, with 84% of executives ranking it as the leading source of value from their AI investments though 10% have also indicated that they have seen an uplift in revenue. What seems to be holding organisations back — and rightly so — are concerns around security, governance, and resilience. Deploying AI responsibly, with the right guardrails and enterprise controls will be the big focus in the next stage of this journey.”

Security and AI governance outrank cost in deployment decisions

Security, risk and governance considerations rank as the single biggest factor shaping how and where companies build and deploy AI systems, ahead of cost, speed and ease of deployment, the study found.

Platform choice is tied closely to that caution. 90% of respondents say model independence matters to their AI investment decisions, and 34% rate AI sovereignty requirements — covering data location, model choice and compliance — as a high or very high organisational risk. That concern runs highest in Europe, where 43% flag it. Despite the worry, only 7% of global respondents have an organisation-wide plan in place to switch AI providers if they need to.

Fragmented knowledge and open-source gains

A few other findings round out the picture. 39% of enterprises are still working out whether their AI initiatives deliver enough value to justify what they are spending on tokens and compute. 44% say important business knowledge is still stuck with individual employees or scattered across documentation, which limits how much context AI systems can actually draw on.

Open-source AI is making headway for specific jobs: 46% of enterprises report targeted use of open-source models for selected use cases. On jobs, leaders are growing less certain rather than more — the share who say it is too early to tell how AI will affect employment rose from 4% in 2024 to 19% in 2026.

Regional patterns differ too. North America scores strongest on enterprise AI foundations, with 15% of leaders there treating AI as a strategic enterprise capability. Europe’s focus is sovereignty, reflected in that 43% risk figure, while the UK and Ireland stand out specifically on operational resilience.