Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s Key Takeaways
- AI agent implementation in businesses is increasingly focused on accountability and governance.
- Approximately half of working hours may undergo transformation due to AI agents.
- Ensuring accountability in AI usage will necessitate that humans are “in the lead” rather than merely “in the loop.”
While agentic AI was seen as largely aspirational in 2025, it has now emerged as a critical performance measure for businesses across various sectors.
Findings from three comprehensive studies conducted by Deloitte, KPMG, and Accenture reveal a disconcerting reality: companies are rapidly adopting AI technology but are lagging in restructuring their operations to effectively integrate human labor with AI agents.
Deloitte’s recent survey on Agentic Transformation indicates that although 43% of organizations are expanding their AI agent deployments across different functions, only 15% have achieved comprehensive, orchestrated multi-agent setups. The survey also highlighted that only 20% of the workforce is deemed ready, and just 16% of businesses feel their processes are adequately prepared for the integration of agentic AI. This stands in stark contrast to the 74% of leaders who anticipate that half of their business processes will be redesigned around AI by 2030.
Also: Businesses must reinvent their processes and workforce to scale agentic AI adoption
Further insights from Salesforce indicate a significant uptick in the number of active AI agents in organizations, which has tripled over the past year. The capabilities of these agents have improved by 350%, allowing them to manage more complex tasks. Additionally, trust in AI agents has led to a threefold increase in employee usage. The average number of agents per organization surged from five to thirteen, with the time required for creation reduced by 53%, averaging just 1.9 days per agent.
From Deployment to Accountability
KPMG’s Global AI Pulse report, based on a survey of 2,145 C-suite executives and business leaders across 20 countries, indicates a significant shift in focus from merely deploying AI agents to ensuring accountability and evaluating the economic impact of AI. This transition is essential as the return on investment from AI adoption remains limited despite increasing deployments.
Also: Business adoption of AI agents tripled this year – as measurable ROI emerges
According to KPMG’s findings, 76% of businesses now recognize tangible value from AI, reflecting a 12% increase over just one quarter. Moreover, 78% of leaders express confidence in their ability to future-proof their AI strategies, marking an 8% rise since Q1 2026. Seventy-one percent report progress towards fully integrating AI and human workforces, up 11% in the same quarter. Organizations that maintain full visibility into AI operating costs are five times more likely to establish a return on investment compared to those lacking such oversight.
While the adoption of AI agents is rapidly increasing, the challenges are also mounting, particularly in scaling use cases and bridging skill gaps, which have emerged as the top barriers to demonstrating ROI.
Key findings suggest that the distinction between organizations effectively deploying AI agents and those struggling is less about the number of agents deployed and more about clear accountability, robust governance, and genuine insight into the costs associated with operating AI at scale.
Also: 12 rules of agentic AI for successful enterprise transformation
Salesforce’s research emphasizes the critical role of leadership in transitioning to agentic business models. Over two-thirds of middle managers express optimism regarding AI’s future role in the workplace and feel a personal obligation to ensure their teams adopt AI tools. The research also indicates that many AI pilot programs prioritize speed and capability while neglecting the essential task of building trust within the organization. The twelve rules for successful agentic business transformation highlight strategies that companies are employing to effectively scale their AI production deployments.
Challenges in Accountability
A joint study by Accenture and Wharton, leveraging Bureau of Labor Statistics task-level data spanning 18 industries, highlights that while “intelligence may be scalable, accountability is not.” Their research indicates that 50% of working hours across the U.S. economy, involving around 120 million workers, are currently influenced by approximately 60 digital and physical AI agents. In the banking and capital markets sectors, digital AI agents alone account for over 45% of hours worked.
Also: ‘Specialists aren’t required’ anymore: How to stay valuable in an AI agent workplace today
Modeling a hypothetical $60 billion company, the study projects potential revenue growth of around $6 billion and annual productivity gains of $1.7 billion from fully matured agentic AI. The report indicates that leading organizations are moving beyond isolated solutions to implement a coordinated array of digital and physical AI agents under human supervision.
The Accenture report reveals that AI agents are proliferating within enterprise systems at a faster rate than governance strategies can adapt. Co-author James Crowley remarked on the importance of having “humans in the lead, not merely in the loop,” underscoring that accountability for tasks must rest with humans, not AI agents. This necessitates a new leadership approach focused on translating expanded capacities into measurable value and sustained growth.
The report also warns that productivity gains will only translate into growth if leaders intentionally redirect newly available capacity towards higher-value tasks; failing to do so may result in stagnation at the efficiency level without achieving growth.
Also: Why replacing staff with AI backfires – and 5 ways smart leaders generate real value instead
Accenture suggests establishing a new business role, the Chief Agentic Resource Officer, alongside explicit profit and loss targets, human-led operational models, and clearly defined decision rights prior to the deployment of AI agents.
Research from Salesforce indicates that 70% of organizations utilizing customer service AI agents achieve a return on investment within 60 days. The adoption rate of agentic AI in service organizations has surged from 39% to 66% in the past year. New outcome-based pricing models could effectively address the accountability challenge by explicitly linking business results to AI agent performance.
Focusing on Relational Transformation
The combined research from Deloitte, Accenture, and KPMG presents a coherent narrative that should reshape how leaders discuss agentic AI. The adoption of AI itself is not the primary challenge; most companies have successfully implemented AI agents, and adoption has increased threefold in the past year.
However, as indicated by Deloitte’s findings on workforce and process readiness, the real challenges lie in the governance frameworks necessary for autonomous AI agents, the gap between deployment and measurable ROI, and the need for disciplined leadership to convert efficiency into growth while maintaining accountability.
Also: The 3 types of people who will excel in the AI agent era, according to tech leaders
None of these firms is advocating against agentic AI; rather, they are converging on a more nuanced argument: the technology has advanced more rapidly than the necessary operating models, governance structures, and workforce readiness to manage it responsibly at scale.
Over the next 12 to 24 months, the distinction will become clear between organizations that see agentic AI adoption as an integration challenge and those that recognize it as a leadership opportunity. Strong human relationships will be essential for navigating the complexities and challenges of transformation.
Business leaders will need to cultivate robust human-AI relationships to ensure that digital labor serves as a source of leverage rather than confusion, passivity, or distrust. They must also consider how the relationships between systems and agents are structured, governed, and monitored.
Becoming an agentic business is fundamentally about relational transformation rather than merely technological change.
About the Author
