The EY 2025 Work Reimagined Survey covered 15,000 employees and 1,500 employers across 29 countries. One finding stood out: 88% of executives reported using AI at work.
Spend a week coaching those executives and a different picture emerges. The vast majority are using AI for basic tasks: rewriting emails, summarizing articles, running simple searches. Surface-level interactions that produce surface-level results.
About 5% are doing something fundamentally different. They use AI to transform how they think, prepare, analyze, and decide. The gap between these two groups has nothing to do with intelligence or technical skill. It comes down to method.
Level 1: The Search Engine
Most CEOs open a blank chat, type a question, and evaluate the answer. Fresh conversation every time. No persistent context. No source materials. No instructions about what a useful answer looks like for their situation.
The results are predictable: generic, surface-level, occasionally wrong. The CEO concludes AI is overhyped for senior-level work and goes back to doing things the old way.
This is like hiring a brilliant consultant and then refusing to brief them. You'd never walk into McKinsey and say "tell me about strategy" without specifying your industry, your competitive position, or your actual question. But that's exactly what a contextless AI prompt does.
Level 2: The Contextual Shift
The first jump is deceptively simple. Tell the AI who you are. Tell it what you're working on. Tell it what constraints you're operating under and what a useful answer looks like for your specific purpose.
One CEO described the difference as going from "Wikipedia answers" to "actually useful." The same model, the same question, dramatically sharper output. The only variable that changed was the input.
At Level 2, AI becomes a capable assistant. It drafts memos that reflect your company's context. It analyzes competitors with knowledge of your positioning. It prepares meeting briefs that account for the specific relationships in the room. Every task improves because the AI has something to work with.
Level 3: The Specialist Team
The 5% operate at a different altitude. They treat AI as a team of configurable specialists, each with its own context, source materials, and instructions.
Think of the briefing room scene in a Mission Impossible film. Ethan Hunt assembles exactly the right experts for the mission: a linguist, a demolitions specialist, a hacker, a disguise artist. Each expert comes pre-loaded with the skills and background needed for the job.
That's Level 3 AI usage. A market intelligence analyst in one project, pre-loaded with your competitor landscape and pricing data. A financial modeler in another, working from your actual spreadsheets. A competitive strategist in a third, briefed on your sector dynamics and your company's specific position.
The Use Cases That Land
Three applications consistently produce visible results in coaching sessions with CEOs.
Sales preparation with real client data. Load a prospect's website, recent earnings calls, and press releases into a workspace. Ask the AI to identify pain points you can address and generate tailored talking points. What used to take an hour of research happens in minutes, with sharper insights because the AI cross-references sources you wouldn't have combined manually.
Competitive intelligence compression. A head of strategy described her competitive analysis cycle as a multi-day process she ran two or three times a month. Same framework, different inputs. In a coaching session, we compressed the data-gathering and cross-referencing steps from days to under an hour. Her judgment still drives the interpretation. The AI handles the assembly.
Document-grounded research. Upload the actual documents, contracts, reports, proposals, regulatory filings, instead of asking AI to recall from its training data. The outputs are grounded in your specific materials, not the model's general knowledge. A family office CEO replaced a $30,000-per-year reporting platform this way.
The Method Gap
The distance from Level 1 to Level 3 is not a technology gap. The tools are the same. The distance is a method gap: knowing how to structure your interaction with AI so the output matches the quality of your thinking.
Most CEOs are stuck at Level 1 because they've never seen Level 3 in action. You can't self-teach a method you don't know exists. And the habits built from twenty years of search engines actively work against the approach that makes AI powerful.
A coaching session built for CEOs compresses months of experimentation into 90 minutes by showing you the levels, live, with your own work. The executives who see Level 3 don't go back to Level 1. The shift is permanent.
If you want to understand what executive AI coaching actually involves, start with the overview.