The Agentic AI Paradox: High Ambition, Low Readiness

A stark disconnect defines enterprise AI in 2026. While 85% of organizations state they want to become agentic within three years, 76% admit their current operations and infrastructure cannot support that change. This is not a technology gap—it is a structural failure. The temptation is to layer AI agents onto existing workflows like a digital bandage. But as Prasun Shah, global CTO for workforce consulting at PwC UK, warns, this is 'adding sticky tapes to parts of an operating model that is breaking.' The result is not transformation but disillusionment.

Why Layering Fails: The Connective Tissue Problem

AI agents are not another application layer. They are connective tissue—moving across systems, coordinating tasks, and making independent decisions. Your existing tech stack was built for human-operated, application-centric workflows. When an AI agent operates at machine speed across multiple systems simultaneously, linear processes break. The solution is not to add more tools but to rewire the operating model. Ema CEO Surojit Chatterjee calls this Agentic Business Transformation (ABT), a framework that rethinks technology, workforce, and metrics as an integrated whole.

Winners and Losers

Winners: Early adopters who redesign their tech stack for AI-native workflows. They will see 30–50% acceleration in business processes and 25–40% reduction in low-value work. Consulting firms like HFS Research and platform providers like Ema will capture premium fees. Losers: Organizations that cling to legacy infrastructure and tool-based metrics. They will watch competitors adapt in days, not months, while their own ROI stagnates. Workers in repetitive roles face displacement unless upskilling programs are launched now.

Second-Order Effects: The Metrics Revolution

When one of Ema’s large customers switched from tool metrics (cost per query, AI accuracy) to outcome metrics (contracts reviewed without human escalation), ROI tripled within two quarters. This is not incremental improvement—it is a paradigm shift. Activity metrics become meaningless when an AI agent handles 1,000 interactions in the time a human handles ten. The new battleground is outcome-based performance management, which will force a complete reconfiguration of reward systems, accountability, and even fiduciary responsibility.

Market Impact: The 2026–2030 Window

McKinsey predicts that by 2030, 75% of current jobs will require redesign, upskilling, or redeployment. The window for action is narrow. Companies that begin ABT now will build adaptive organizations that can configure AI employees in days, not months. Those that delay will face a widening gap between ambition and execution, losing talent, market share, and relevance.



Source: MIT Tech Review AI

Rate the Intelligence Signal

Intelligence FAQ

ABT is a framework coined by Ema and HFS Research that rethinks technology stack, workforce design, and success metrics as an integrated system for deploying AI agents, rather than layering them onto existing operations.

Their current infrastructure and operating models are designed for human-led, application-centric workflows. AI agents require machine-speed, cross-system connectivity and outcome-based metrics that legacy systems cannot provide.

By adopting ABT: redesigning the tech stack to enable AI agents as connective tissue, redefining workforce roles and management, and shifting from activity metrics to outcome-based performance measures.