Zahi Abdein

Briefing · September 29, 2026

AI performance is an operating model problem, not a talent problem

By Zahi Abdein

Or share via

Top story

Only 19% of AI users work in conditions where their organization actively enables the value of AI, according to Microsoft's 2026 Work Trend Index, which surveyed 20,000 AI users across 10 countries. Organizational culture, manager support, and talent practices account for more than twice the reported AI impact of individual mindset and effort (67% versus 32%). The remaining 81% of AI users are either stalled, blocked by organizational systems that haven't caught up, or still emerging. For boards and C-suite leaders, this is the clearest data yet that AI performance is a systems design problem at the organizational level, not a talent or technology problem.

Quick hits

  • Active agents in the Microsoft 365 ecosystem grew 15x year over year, 18x in large enterprises. Yet only 13% of AI users say they are rewarded for reinventing work with AI even when immediate results are not met. The incentive system is working against the transformation organizations say they want. Source: Microsoft Work Trend Index 2026
  • Most EU AI Act obligations, including transparency duties and rules for high-risk AI systems, came into force on August 2, 2026. Board-level AI governance disclosures in public company filings have grown 84%. Forrester predicts 60% of Fortune 100 companies will appoint a dedicated head of AI governance in 2026. Source: EU AI Act / Forrester 2026
  • 76% of firms now have a Chief AI Officer, up from 26% in 2025, a 50-point jump in one year. Sony, Bank of America, and UBS have formalized board-level AI governance structures. The structural shift from AI as a technology function to AI as a governance responsibility is now largely complete - execution remains the challenge. Source: CNBC / Forrester 2026

Insight for practice

The data this week reinforces a consistent pattern: AI returns are a systems design problem. When only 19% of AI users work in environments where organizational conditions reinforce their capability, and organizational factors drive twice the impact of individual effort, the strategic intervention point is clear. The most useful diagnostic question for boards and C-suite advisors is not 'are our people skilled?' but 'does our operating model reward, support, and amplify AI-enabled work?' The organizations winning with AI are redesigning their culture, management practices, and talent systems in parallel with their technology. That is a board and leadership agenda, not an IT one.

Worth reading