Skip to main content

The Research

88% of organizations use AI.
81% see no real return.

This is the gap. It is not a technology problem. It is a thinking problem — and it has a specific, researchable cause.

McKinsey State of Organizations (2026)

88%

of organizations now use AI in at least one function

81%

report no significant bottom-line impact from their AI investments

21%

have redesigned workflows around AI — 2.8× more likely to see real EBIT impact

McKinsey State of Organizations (2026)

Why the gap exists

Organizations add AI to existing work.
The 21% redesign the work itself.

The factor most correlated with meaningful AI returns is not budget, technology choice, or sector. It is workflow redesign: examining how work actually happens and rebuilding it around what AI can now do reliably.

The other 79% add AI to existing processes and measure the result in disappointment. The system stays the same. The system produces the same results, slightly faster.

The investment ratio that changes everything

$1Tech
$5People

For every dollar spent on AI technology, winning organizations invest five in people: judgment, strategic thinking, and the capacity to use AI well.

McKinsey State of Organizations, 2026

Automation Exposure by Skill Type

Not all skills are equally at risk.

McKinsey mapped 800+ occupations and 2,100+ activities. The pattern is consistent: physical and cognitive execution is most exposed; social, creative, and judgment-based work is least.

High automation exposure
Data collection & processing81%
Predictable physical work78%
Routine cognitive tasks69%
Applied expertise (knowledge work)28%
Low automation exposure
Stakeholder interaction20%
Managing & developing people9%
Creative problem-solving6%

MGI, "Agents, Robots, and Us," 2026

How Work Evolves

Same role. Different work.

Professional work migrates upward — from execution to interpretation, from preparation to judgment. AI handles the preparation; you handle the meaning.

Before AI integration
After deliberate integration
Hours of data analysis
Minutes of analysis → hours of interpretation & decision-makingAI compiles and patterns; you contextualize and decide
Research compilation & document prep
AI first drafts → your question-framing & quality controlThe bottleneck shifts from finding information to asking the right questions
Demand forecasting from spreadsheets
AI-generated forecasts → scenario planning & strategic betsPrediction is automatable; wisdom about which prediction matters is not
Routine client updates
AI handles updates → you focus on relationship & trust-buildingOne-third of work relies on social-emotional capacities AI can't replicate
Learning new tools every quarter
Building transferable judgment that works across any tool"Transferable skills matter more than specialization when circumstances shift rapidly" — MGI

MGI, "Agents, Robots, and Us," 2026

The Value Migration

Your job title stays.
Your actual work transforms.

Professional work migrates upward. AI handles the preparation; you handle the judgment.

Hours of data analysis

Minutes of analysis → hours of interpretation & decision-making

AI compiles and patterns; you contextualize and decide

Research compilation & document prep

AI first drafts → your question-framing & quality control

The bottleneck shifts from finding information to asking the right questions

Demand forecasting from spreadsheets

AI-generated forecasts → scenario planning & strategic bets

Prediction is automatable; wisdom about which prediction matters is not

Routine client updates

AI handles updates → you focus on relationship & trust-building

One-third of work relies on social-emotional capacities AI can't replicate

Learning new tools every quarter

Building transferable judgment that works across any tool

"Transferable skills matter more than specialization when circumstances shift rapidly" — MGI

MGI, "Agents, Robots, and Us," 2026

800+ Occupations Mapped

Where does your work sit?

McKinsey segments all occupations by which partner — people, agents, or robots — handles the majority of tasks. This determines your disruption exposure and your opportunity.

People-Centric

Minimal Disruption

Nurses, therapists, coaches, social workers, firefighters

Disruption exposure15%
People + Agents

Evolving Roles

Sales, HR, educators, consultants, project managers

Disruption exposure40%
Agent-Centric

Significant Shift

Accountants, lawyers, developers, analysts, writers

Disruption exposure70%
People + Robots

Physical Evolution

Electricians, plumbers, tradespeople, technicians

Disruption exposure30%
Robot-Centric

High Automation

Logistics workers, welders, assembly line, warehouse

Disruption exposure75%
Hybrid

All-Three Roles

Medical assistants, receptionists, operations coordinators

Disruption exposure50%

MGI, "Agents, Robots, and Us," 2026

The Partnership Model

People. Agents. Robots.

McKinsey maps 800+ occupations into a three-way partnership. The shift isn't replacement — it's redistribution of what each partner does best.

McKinsey maps every occupation into a primary partnership model — the mix of humans, AI agents, and physical robots that handles the majority of its tasks. The $2.9 trillion figure represents the combined economic value unlocked by 2030 when all three partners operate at full capacity.

People — 43%

Social interaction, ethical judgment, creative direction, leadership. The work that requires genuine human presence and accountability.

AI Agents — 44%

Knowledge synthesis, research, drafting, analysis, coordination. Software agents handling cognitive tasks at scale.

Robots — 13%

Physical logistics, manufacturing, warehouse operations. Robotics still lack dexterity for complex physical environments.

People 43%
AI Agents 44%
Robots 13%

43%

People

Social interaction, ethical judgment, creative direction, leadership. The work that requires genuine human presence and accountability.

44%

AI Agents

Knowledge synthesis, research, drafting, analysis, coordination. Software agents handling cognitive tasks at scale.

13%

Robots

Physical logistics, manufacturing, warehouse operations. Robotics still lack dexterity for complex physical environments.

Over 70% of skills employers seek today apply to both automatable and non-automatable work — they evolve, they don't disappear. MGI, "Agents, Robots, and Us," 2026

The Fastest-Growing Skill

in two years.

AI fluency demand in global job postings grew from roughly 1 million to 7 million in just two years — faster than any professional skill on record. But fluency is not prompting. It is judgment.

AI fluency is not

Knowing every prompt trick. Chasing new tools. Replacing your judgment with AI output. Speed without direction.

AI fluency is

Knowing when to use AI and when not to. Framing the right questions. Interpreting results with domain expertise. Deliberate judgment.

AI fluency in global job postings
~1M postings
2023
~3M postings
2024
~7M postings
2025

McKinsey Global Institute, 2025–2026

The Netherlands, 2030

More jobs. Fewer prepared people.

The Netherlands isn't losing jobs — it's reshuffling them. A net gain of 250,000 roles by 2030. But only for those who make the shift intentionally.

Net +250,000 jobs
+400KCreated
−150KDisplaced

McKinsey Global Institute, "How AI Reshapes Work and Skills in Europe," 2026

+400KJobs created in growing segments
−150KJobs displaced in declining segments
15%of FTE hours automatable — below EU average
Labor market tightness tripling by 2030

The question isn't whether this shift happens. It's whether you're positioned for the 400,000 or the 150,000.

AI exposure by sector

Finance & Insurance68%
Information Technology61%
Public Administration54%
Education47%
Professional Services43%

% FTE hours automatable — EU comparison

Germany
18%
France
16%
Netherlands
15%
Sweden
14%
Poland
11%
Spain
10%

Cross the gap

The gap is not fixed. It is a choice.

Whether you are an individual professional or an organization, the path from the 79% to the 21% is the same: examine how work actually happens, and be willing to change it.