Genpact and HFS Research have released a new study, Autonomy Requires Trust in AI, highlighting a critical disconnect between the ambition for agentic AI and current organizational readiness. While 92% of senior executives believe autonomous AI systems will fundamentally reshape work, nearly 80% still keep these systems in supervised modes, requiring human approval for most actions. The report identifies that the transition from assistive productivity tools to autonomous execution layers depends on four pillars: accountability, measurement, people, and process.
Market Sentiment: 92% of leaders believe agentic AI will fundamentally change work execution.
The Autonomy Paradox: 80% of systems are still human-supervised; only 22% of firms are comfortable with broad autonomy.
Investment Surge: Spending on agentic AI is projected to rise by 38% in the next year.
Organizational Shift: 44% of executives anticipate flatter management structures as AI handles coordination.
Top Barriers: Regulatory risk, lack of explainability, and unprepared business processes (cited by 33%).
Skill Evolution: High demand for workflow orchestration (42%), data engineering (39%), and monitoring (36%).
The research, based on a survey of 545 senior executives across 11 industries, suggests that scaling agentic AI is no longer a technical challenge but an operational one. Enterprises have successfully used Generative AI to accelerate tasks; however, the next phase requires AI to take responsibility for executing workflows. Currently, the "trust gap"—driven by concerns over reputational risk and accountability—remains the primary reason agents are not granted more independence.
"Enterprises have proven that generative AI can accelerate work; the next frontier is proving that agentic AI can take responsibility for executing it," said Ajay Vasal, Global Leader for Data & AI, Genpact. "Success depends on how organizations redesign processes with accountability and decision rights built in."
The report outlines that organizations must move beyond outdated productivity metrics, which 67% of firms still rely on, and adopt "agent-native" KPIs. As AI absorbs coordination work, management layers are expected to decrease, requiring a significant rethink of human roles. The focus is shifting from building AI models to orchestrating and observing them in production.
Phil Fersht, CEO and Chief Analyst of HFS Research, noted, "Ambition for autonomy is outpacing the ability to govern it. The leaders in agentic AI won’t be the fastest adopters, they’ll be the ones who redesign accountability, measurement, and human oversight."
To bridge the gap and move toward autonomous execution, the report recommends four critical actions:
Accountability: Clearly defining ownership for an agent’s actions and potential failures.
Measurement: Tracking autonomous workflow completion rather than just individual task speed.
People: Providing role clarity for humans whose primary job will shift to oversight and intent-setting.
Process: Eliminating manual handoffs and redesigning workflows for a digital-first execution model.
The study concludes that while the technology for agentic AI is rapidly maturing, the enterprise "operating system"—the processes and governance frameworks—must undergo a fundamental redesign to unlock the projected value of autonomous labor.
ABOUT GENPACT
Genpact (NYSE: G) is an agentic and advanced technology solutions company. We leverage process intelligence and artificial intelligence to deliver measurable outcomes. With a strong partner ecosystem and decades of client trust, we provide innovative solutions that transform how businesses run. Powered by a team with an active learning mindset and client centricity at its core, we deliver lasting value for the world’s leading enterprises.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.