Rethinking organizational design in the age of agentic AI
View original at technologyreview.comMIT Technology Review - Ai Research Title: Rethinking organizational design in the age of agentic AI Date: 2026-05-26 14:54 Source: https://www.technologyreview.com/2026/05/26/1137584/rethinking-organizational-design-in-the-age-of-agentic-ai/ <p>Amid rapidly growing adoption of enterprise-level AI agents, there’s a dis…
What we drew from this source
The claims Via News extracted from this document. We point to the source; we don't replace it.
Existing enterprise technology stacks designed for human-operated, application-centric workflows must be reconsidered when the actor is an AI agent operating at machine speed across multiple systems simultaneously
60% confidenceAI agents derive their value not as another layer in a technology stack, but as connective tissue moving across layers to coordinate tasks and contextualize data from multiple applications—this is the next competitive battleground for enterprises
60% confidenceExisting vocabulary—digital transformation, AI transformation, co-pilot—fails to capture the full scope of AI agent-driven organizational change; ABT is categorically different as it represents integration of AI agents into the fabric of the organization
60% confidenceManagers in hybrid human-AI workforces will be freed from execution-based tasks but must manage new tensions around trust, explainability, psychological safety, and status dynamics within hybrid teams
60% confidenceActivity-based workforce metrics become meaningless or actively misleading when AI employees are introduced; an AI can handle a thousand customer interactions in the time a human handles ten, masking whether those interactions drove customer satisfaction, retention, or revenue
60% confidenceOrganizations that make the architectural shift to agentic AI can configure AI employees using natural language, compressing the time from business requirement to production workflow from months to days
60% confidenceThe ABT framework drives the need to redesign an organization in its entirety—operating model, workflows, decision rights, and performance management systems—ensuring AI agents are active participants in value creation rather than point tools
60% confidenceOrganizations are embedding AI employees into a human operating model by layering AI agents onto existing workplace structures rather than reimagining the operating model; this is equivalent to adding sticky tape to a breaking system
60% confidenceBy 2030, three-quarters of current jobs will require redesign, upskilling, or redeployment; organizations must act swiftly to amend recruitment, retention, and remuneration policies
60% confidence85% of organizations say they want to be agentic within the next three years, but 76% say their current operations and infrastructure cannot support that change, citing lack of readiness across people, processes, and workflows
60% confidenceAI agents could accelerate business processes by 30-50% and reduce low-value work time by 25-40% when deployed at scale across customer service, HR, and sales
60% confidenceIn human-AI teams, operational accountability will become significantly more diffused to reflect AI agents' systemic role, while ethical and fiduciary responsibilities will likely remain with human employees; senior leaders must determine accountability when AI makes mistakes and what guardrails protect customers
60% confidenceAn Ema enterprise customer tripled measured ROI from agentic AI within two quarters after shifting from tool metrics to outcome metrics, pivoting from point solutions in high-volume, low-complexity workflows to deploying AI employees where outcome value was highest
60% confidence
Cited in these Via News reports
- 88% of Enterprise AI Leaders Globally Regret Skipping Foundational Work as Agentic Scale Arrives →
- 88% of Global Executives Regret Weak AI Foundations as Agentic Infrastructure Goes Live →
- From Co-Pilots to Autonomous Agents: Snowflake, NVIDIA, and Dell Build the Global AI Infrastructure Layer →
- GPU Lead Times Reach 52 Weeks as Global Enterprise AI Race Moves Beyond Copilots to Autonomous Agents →
- Global Enterprises Integrate AI Agents as Autonomous Workforce Members →
- Global Enterprises Replace AI Assistants With Autonomous Agents, Forcing Deep Operational Restructuring →
- Microsoft, Snowflake Expand AI Agent Platforms as Governance Rules Lag →
- Snowflake, EXL, and Netskope Ship Agent Infrastructure as Global Enterprise Tech Stacks Face Replacement →