The State of AI – How Organizations Are Rewiring to Capture Value (McKinsey Executive Brief)

Executive Summary

This executive document summarizes McKinsey's 2025 report on the state of AI and gen AI. It explains how organizations are reorganizing their strategies, architectures, data, and operating models to move from experimentation to enterprise-scale value capture.

Original article:  The state of AI: How organizations are rewiring to capture value »

Section I

From Experiments to Rewiring the Enterprise

McKinsey’s survey shows that AI has moved from isolated pilots to a broad transformation that touches strategy, technology, data, and culture. Organizations are “rewiring” their core to embed AI into how decisions are made and work gets done.

Key patterns

• AI usage spans multiple functions: marketing, sales, operations, supply chain, risk, HR, and software.
• Leaders move away from one-off PoCs toward a portfolio of AI “engines” that serve the entire organization.
• Gen AI lowers the barrier to experimentation, but raises the bar on governance and integration.

Section II

Gen AI Surge: Adoption, Use Cases & the Value Gap

The survey finds that organizations are most often using gen AI in marketing and sales, product and service development, customer service operations, software engineering, IT, and knowledge management. These are domains with rich unstructured data and clear opportunities for copilots and content generation.

Typical use patterns

• Marketing & sales: content generation, campaign optimization, personalization.
• Product development: idea generation, requirement synthesis, design exploration.
• Service operations: chatbots, email triage, knowledge retrieval for agents.
• Software engineering: code synthesis, documentation, test generation.
• Knowledge management: summarizing and surfacing enterprise knowledge.

A central tension of the report is that, although gen AI adoption is high, the majority of organizations do not yet see a significant contribution to enterprise-level earnings from these initiatives. Revenue increases and cost reductions are visible within specific functions, but not yet at full-enterprise scale.

Why the value gap persists

• The focus is often on experiments and demos rather than systemic workflow redesign.
• Gen AI is deployed in pockets, without unified platforms, data, and governance.
• Adoption practices (KPIs, change management, incentives) are immature in most organizations.

Section III

Rewiring the Organization: Strategy, Platforms, Data & Talent

The report identifies a subset of “AI high performers” that achieve significantly greater impact from AI and gen AI. These organizations follow more of the known adoption and scaling best practices and demonstrate stronger governance and technical foundations.

What high performers tend to do

• Track clear, value-linked KPIs for AI and gen AI solutions across functions.
• Establish a unified roadmap and centralized AI governance, often with CEO oversight.
• Deploy AI platforms that can serve multiple use cases instead of building one-off solutions.
• Reskill employees at scale and design incentives that reinforce AI adoption.
• Mitigate more AI-related risks proactively (cybersecurity, IP, privacy, inaccuracy).

Section IV

Risk, TRiSM & Organizational Challenges

Organizations are increasingly investing in mitigating gen-AI-related risks. The most commonly managed risks include inaccuracy of outputs, cybersecurity threats, intellectual property infringement, data privacy, and regulatory compliance. Larger organizations are more active in risk mitigation, but even these are still maturing their practices.

Key TRiSM themes

• Policy and guidelines for responsible gen AI use across the enterprise.
• Controls around access to models, data, and tools; logging and monitoring.
• Human review of gen AI content in high-stakes or external-facing contexts.
• Emerging focus on explainability and workforce impacts, not just security.

Many of the barriers to value are organizational rather than technical: unclear ownership, fragmented pilots, limited training, insufficient incentives, and lack of trust in AI outputs. McKinsey emphasizes that scaled impact requires robust change management and leadership engagement.

Typical frictions

• Lack of clear accountability for AI products after launch.
• Teams unsure how to integrate AI into existing workflows and decision rights.
• Underinvestment in training and enablement for non-technical users.
• Concerns about job impact, quality, and fairness of AI decisions.

Section IV

Risk, TRiSM & Organizational Challenges

Organizations are increasingly investing in mitigating gen-AI-related risks. The most commonly managed risks include inaccuracy of outputs, cybersecurity threats, intellectual property infringement, data privacy, and regulatory compliance. Larger organizations are more active in risk mitigation, but even these are still maturing their practices.

Key TRiSM themes

• Policy and guidelines for responsible gen AI use across the enterprise.
• Controls around access to models, data, and tools; logging and monitoring.
• Human review of gen AI content in high-stakes or external-facing contexts.
• Emerging focus on explainability and workforce impacts, not just security.

Many of the barriers to value are organizational rather than technical: unclear ownership, fragmented pilots, limited training, insufficient incentives, and lack of trust in AI outputs. McKinsey emphasizes that scaled impact requires robust change management and leadership engagement.

Typical frictions

• Lack of clear accountability for AI products after launch.
• Teams unsure how to integrate AI into existing workflows and decision rights.
• Underinvestment in training and enablement for non-technical users.
• Concerns about job impact, quality, and fairness of AI decisions.