AI agents could replace org charts, Microsoft’s Asha Sharma says

Introduction

On August 28, 2025, Microsoft AI Platform CVP Asha Sharma said on Lenny’s Podcast that “the org chart starts to become the work chart,” arguing AI agents will push companies toward task-based structures with fewer layers. Speaking about how agents route work, she added that employees may bring personal “agent stacks” to expand their skills and output.

Why It Matters

Signals a structural shift from role-based hierarchy to throughput-driven, task networks.
Could flatten management, change spans of control, and alter promotion paths.
Sets expectations for “bring-your-own agent” practices, governance, and security.

Details / Specs / Numbers

Core thesis: AI agents embedded in workflows will coordinate tasks and escalate to humans as needed; throughput trumps hierarchy.
“Work chart” vs. org chart: Teams organize around loops and tasks, not reporting lines; fewer management layers.
Agent stack: Employees could bring a curated set of AI tools/agents to augment skills—akin to BYOD but for capabilities.
Adoption context: Big Tech has been delayering; analysts and executives cite efficiency and bureaucracy reduction as drivers.
Recent figures: Microsoft cut ~6,000 roles (≈3%) in May 2025 and later signaled further flattening; Google told staff it reduced managers of very small teams by ~35% over the past year.

Timeline & Official Statements

August 28, 2025 — Lenny’s Podcast publishes Sharma’s episode discussing “the death of the org chart” and the rise of agent-driven “work charts.”
May 13, 2025 — Reports indicate Microsoft is cutting ~6,000 roles (~3%) to reduce management layers.
July 2, 2025 — Microsoft outlines further streamlining and fewer organizational layers amid heavy AI investment.
August 27–29, 2025 — Reports from multiple outlets cite Google executives saying the company cut ~35% of managers overseeing very small teams to reduce bureaucracy.

Market/Industry Impact

Flattening organizations may speed decisions and reduce handoffs, but raises questions about oversight, accountability, and AI safety. HR leaders will need updated job architectures, evaluation criteria for agent-augmented output, and clear guardrails (data access, prompt security, model choice). For vendors, the shift favors platform-level agents integrated with productivity suites, while consulting/services may surge around agent orchestration, change management, and compliance.

What to Watch Next

Enterprise policies for BYO agent stacks (security, licensing, data governance).
Tooling for agent routing and observability (task assignment, audit trails, RLHF/feedback loops).
Changes to manager spans of control and revised career ladders for ICs vs. managers.
Evidence from pilots linking agent adoption to measurable productivity and quality gains.

TL;DR

Microsoft’s Asha Sharma says AI agents could replace org charts, shifting firms to task-based “work charts.”
Big Tech is already delayering; companies report fewer managers and wider spans of control.
Expect policies for “bring-your-own agent stacks,” plus new governance and measurement tools.

FAQ

Q: What does “work chart” mean?
A: A dynamic map of tasks and throughput—who/what agent handles which step—rather than static reporting lines.

Q: Does this predict mass layoffs?
A: Sharma describes fewer layers and broader skill sets, not necessarily fewer jobs; roles and workflows may change as agents handle routine work.

Q: What is an “agent stack”?
A: A personal set of AI agents/tools an employee uses to extend capabilities (e.g., research, coding, scheduling), subject to company policy.