TCS Study Reveals 84% of Executives Cite Productivity Gains from AI but Only 10% Allow Autonomous Agents

Brickinfo News Agency – A global study conducted across 25 countries by Tata Consultancy Services (TCS) reveals that 84% of surveyed executives consider productivity improvement the primary source of value from artificial intelligence (AI) investments, yet only 10% permit AI agents to execute decisions without human oversight across business functions.

The report, titled “TCS Enterprise AI Study: Intelligence that is Real,” surveyed 1,333 senior business and technology leaders across 12 industries. The findings show that security, risk, and governance have become the primary considerations shaping AI development and deployment, ranking ahead of cost, speed, and deployment convenience. Furthermore, only 9% of executives reported that their AI agents can securely orchestrate workflows across enterprise systems, demonstrating an operational gap between AI-driven assistance and fully autonomous execution.

Amit Kapur, Chief AI and Services Transformation Officer at TCS, stated that enterprise AI is transitioning from task assistance to autonomous operations with deeper systems integration. He noted that while 84% of organizations derive primary value from productivity and 10% report revenue gains, security, governance, and operational resilience remain the main factors preventing companies from granting AI broader autonomy.

The research also identifies rising attention toward platform resilience and AI sovereignty. Approximately 90% of respondents identified model independence as a significant factor in investment decisions, while 34% classified sovereignty requirements—such as data location, model selection, and regulatory compliance—as a high or very high organizational risk. In Europe, concern over AI sovereignty reached 43%. However, only 7% of organizations globally have implemented company-wide provider-switching continuity plans.

Additional findings indicate that 39% of organizations are assessing whether project value justifies compute and token costs, and 44% report that institutional knowledge remains trapped in unstructured formats or employee silos. Meanwhile, 46% have adopted open-source AI for selected functions, and the percentage of leaders who say it is too early to measure the impact of AI on employment increased from 4% in 2024 to 19% in 2026.

The full report, TCS Enterprise AI Study: Intelligence that is Real, is available at: TCS Enterprise AI Study: Intelligence that is real