The report describes a **“new agency equation”** where AI and agents take on more of the execution work, while humans gain more room to **direct, decide, and own outcomes**. Instead of replacing human judgment, AI is increasingly used to **support cognitive work**—things like analyzing information, solving problems, evaluating options, and thinking creatively. Some key points from the research:

• A privacy-preserving analysis of more than 100,000 Microsoft 365 Copilot chats found that 49% of all conversations involved cognitive work (analysis, problem-solving, evaluation, creative thinking).
• 58% of AI users say they’re producing work they couldn’t have a year ago; this rises to 80% among “Frontier Professionals” (those using AI in more advanced ways).
• Most AI users treat AI output as a starting point, not a final answer, and stay responsible for the thinking and decision-making.

In practice, this means AI and agents handle more of the execution and information processing, while people focus on **direction-setting, quality control, and critical decisions**. The opportunity for human potential expands—but only if organizations are structured to capture it.

The data shows that the main driver of AI impact is **institutional, not individual**. According to the report:

• Organizational factors—such as culture, manager support, and talent practices—account for about 67% of reported AI impact.
• Individual factors—like personal mindset and behavior—account for only about 32%.

This means that even if employees are motivated and capable with AI, their impact is limited if:

• Managers don’t actively support or model AI use.
• Goals, incentives, and processes are still designed for pre-AI ways of working.
• The organization doesn’t capture and share what’s being learned from AI in day-to-day work.

By contrast, “Frontier Firms” are pulling ahead because they:

• Rearchitect work—clarifying what humans do and what AI does.
• Encourage managers to model AI use, which creates a measurable lift in recognized AI value.
• Treat the company as a Learning System, where outputs and insights from AI-supported work are captured and reused.

For your organization, this suggests focusing less on one-off AI training and more on **culture, leadership behavior, and talent systems** that normalize and reward new ways of working with AI.

The **“Transformation Paradox”** describes a tension many organizations are feeling as AI adoption accelerates. The research highlights a few conflicting pressures:

• 65% of AI users fear they’ll fall behind if they don’t adopt AI quickly.
• At the same time, 45% say it feels safer to stick to current goals than to redesign how work gets done.
• Only 13% feel rewarded for reinventing how they work.

So people feel pressure to use AI, but the systems around them often reward short-term performance over experimentation and change. Practical responses include:

• Make reinvention part of the job: Explicitly recognize and reward teams that rethink workflows with AI, not just those who hit existing targets.
• Support human agency: Emphasize skills like quality control of AI output and critical thinking—identified by AI users as the top human skills as AI takes on more work (50% and 46% respectively).
• Build a Learning System: Treat every AI-assisted project as a source of insight. Capture what worked, share it, and fold it back into how the organization operates.

By addressing the paradox directly—balancing performance expectations with space to reimagine work—organizations can move from simply adopting AI tools to **reshaping how work gets done** in a sustainable way.