The Leader Is the Strategy

¡AY AY AY, AI!

Weekly newsletter to help you master the new rules for success in the AI age

My grandfather was a rancher. He worked land in Alpine and in Presidio, Texas, two of the most unforgiving stretches of West Texas you can find. He had people he trusted. He had help. But he was always watching.

He knew which horse or cattle was off before anyone told him. He felt when the water was running low before it became a problem. I never once saw him fully step back and say, handle it. He was present because he understood something that took me years to articulate: the ranch ran on his attention, not just his investment.

He had a saying he came back to again and again.

El ojo del amo engorda al caballo.
The master's watchful eye is what makes the horse grow fat.

He did not mean a rancher had to do everything himself. He meant no one cares about the outcome the way the owner does. That presence, that personal accountability, that is the variable.

I thought about him a lot this week. Because new data just dropped that proves his ranch wisdom holds in the boardroom too.

KPMG released its Q2 2026 Global AI Pulse, surveying more than 2,000 business leaders across 20 countries. The headline finding: when the CEO personally owns accountability for AI, 57% of organizations report meaningful business value. When no one at the top owns it, that number falls to 21%.

We see the same pattern in our own work. The most senior leaders from some of the most respected Fortune 50 and Fortune 100 companies who commit personally to reinventing themselves as agentic leaders do not just change themselves. They bring their whole teams with them. Our own FlipFactor™ data consistently show that when a leader makes a genuine personal commitment to becoming an agentic human, their teams shift. Entire groups move from the guardian and explorer archetypes into the orchestrator quadrant. The whole team moves, not because the leader issued a mandate, but because the leader modeled the way.

The KPMG data confirmed what we have been watching happen in real time. The tool did not change. The technology did not change. The variable is the leader.

My grandfather would have called that obvious.

THIS WEEK'S INSIGHTS

The question this week is not which AI tool you are using. It is who in your organization is close enough to make it work.

1. CEO accountability is the single largest predictor of AI ROI, and the gap is not small. KPMG's Q2 2026 Global AI Pulse of more than 2,000 leaders across 20 countries found that when the CEO personally owns AI strategy, 57% of organizations report meaningful business value. When ownership is unclear, that number is 21%. Established ROI more than triples, from 4% to 14%. This is not a rounding error. This is the difference between transformation and an expensive subscription.

Think of it like a construction project. You can hire the best crew and buy the best materials, but if no one with real authority is walking the site, the work drifts. Every contractor knows this. So does every rancher.

2. Most companies are investing in AI, but very few have anything to show for it yet. Only 7% of organizations in the KPMG survey report established ROI, even as 79% name AI a top investment priority. And leaders with strong visibility into what their AI actually costs are 5x more likely to achieve that ROI. The gap is not a technology failure. It is an accountability and attention failure. You cannot run a ranch by just paying the feed bill.

3. The companies making real progress are investing in people, not just platforms. Human-AI collaboration jumped from 60% to 71% quarter over quarter in the KPMG data. And 48% of organizations are actively upskilling their workforce for the agentic era, with adaptability now outranking technical skills as the most valued attribute for new roles. The organizations moving fastest are not buying more tools. They are building more capable people to direct them.

3 MYTHS TO REFRAME

Myth #1: "We have an AI strategy."

Why we believe it: Leadership approved a budget. Someone in IT manages the subscriptions. There is a roadmap somewhere. That counts.

Reframe: Most organizations have AI activity, not AI strategy. The KPMG data draws a sharp line between the two. Activity looks like spending and adoption. Strategy looks like named ownership, clear accountability, and a leader who personally tracks whether any of it is working. Only 7% of organizations report established ROI, while 79% say AI is a top priority. That gap does not close through better tools. It closes through better accountability.

What to do: Ask one question in your next leadership meeting: Who is personally accountable for AI results in this organization? Not who approved the budget. Who owns the outcome? If it takes more than five seconds to answer, you have found your gap.

Myth #2: "AI decisions belong to someone on my team."

Why we believe it: There are people smarter than me about this technology. I hired them. This is their lane.

Reframe: The moment AI agents start routing decisions, automating workflows, and changing how work gets done across your organization, you no longer have a technology project. You have a redesign of how the firm operates. Those are not decisions you can fully delegate. The KPMG data makes the cost of delegation visible: when the CEO is not personally accountable, meaningful business value drops from 57% to 21%. Your team can execute. Only you can own.

What to do: Pick one AI initiative this week. Ask whether you personally understand what it is doing, what it costs, and what result it is supposed to produce. If the honest answer is no, that is your starting point, not someone else's.

Myth #3: "If my team is resisting AI, they will come around eventually."

Why we believe it: Resistance is normal with new technology. Give it a cycle, and people adapt. This is not urgent.

Reframe: Resistance does not dissolve on its own. The KPMG data show that the top two drivers of AI agent resistance among employees are concerns about increased workload (58%) and fears about job security (58%). Neither of those resolves without a direct conversation. The resistance is not about the tool. It is about what the tool means for people's roles, and no one has explained that clearly enough yet.

What to do: Pick one person on your team who has gone quiet on AI. Ask them directly: "What part of your work feels most uncertain when you think about how we're using AI?" Listen before you answer. Then tell them what AI will and will not touch in their role. Clarity, even imperfect clarity, beats silence every time.

TOOLS TO EXPLORE

This week, try the following tools and Prompts to Steal.

Claude as a thinking partner for accountability mapping

Prompt to steal:

"I want to understand where AI is being used across my team and who owns each initiative. Ask me about my major workflows one at a time. After each one, note who I named as accountable for the outcome. When we've covered them all, show me which workflows have clear ownership and which ones don't. Ask me follow-up questions before you summarize anything."

Power tip: Do this exercise yourself before you ask your team. The gaps you notice in your own answers are usually the same gaps your team is navigating without telling you.

NotebookLM to audit your AI progress

Prompt to steal:

"Upload the last three AI initiative updates from my organization. For each one: what result was promised, what was actually measured, who was accountable, and whether any result was verified by someone outside the team doing the work. Stay inside these documents only. Show me where the accountability trail breaks down."

Power tip: NotebookLM only answers from what you upload. That constraint is the point. It shows you what your documentation actually says, not what you assumed it said.

TRY IT THIS WEEK (Micro Actions)

1. Name your AI owner.

If you lead a team: Find the most significant AI initiative running in your organization right now. Write down one name: the person personally accountable for that initiative's business results. Not the vendor relationship. Not the IT lead. The person who will be asked to explain in six months why it worked or why it did not. If you cannot write down a name, that is your action item.

If you are building your career: Ask your manager or a senior leader one direct question this week: "Who owns the outcomes for our AI initiatives?" You are not being difficult. You are thinking like a strategist. If they cannot answer quickly, you have just identified a gap that matters. Pay attention to how they respond. That tells you a lot about where your organization actually is.

2. Have the job security conversation.

If you lead a team: Pick one person on your team who has gone quiet on AI. Ask them directly: "What part of your work feels most uncertain when you think about how we're using AI?" Then listen. Do not explain or defend. Just listen first. You will learn more from that one conversation than from any survey.

If you are building your career: Have this conversation with yourself first. Write down the parts of your role that feel most exposed as AI expands. Then proactively bring that question to your leader. Do not wait for someone to address it. The professionals who get ahead in this moment are the ones who name the uncertainty out loud and then act on it. That starts with you.

3. Use the tool yourself for 30 minutes.

If you lead a team: Not reviewing someone else's AI output. Actually complete a task you normally do, using an AI agent, from start to finish. This is not about efficiency. It is about staying close enough to the technology to lead it from real knowledge instead of secondhand reports.

If you are building your career: Pick a task you do regularly and run it two ways: once with AI, once without. Compare the results. Notice where your judgment added something the AI could not. That gap is your value. Understanding it clearly is how you position yourself, not as someone AI replaces, but as someone who directs it.

POWER TIP

Before your next AI budget conversation, answer three questions first: Who owns this outcome? How are we measuring it? Who reviews the results and has the authority to change course? If those answers reside in a deck rather than in a person's job description, the investment is exposed. Get ownership visible before you approve the next dollar.

👉 Who in your organization personally owns your AI results? And do they know it?

Closing Thought

The companies making AI work are not the ones with the biggest budgets or the most tools. They are the ones with a leader who stays in it, asks hard questions, and refuses to watch from a distance.

Satya Nadella recently wrote something that landed for me. His essay, which has been read more than 65 million times, put it plainly: your company does not need an AI strategy. It needs a learning loop you own. The companies that win are not the ones that picked the best model. They are the ones turning their workflows, their human judgment, and their corrections into a compounding system over time. Everyone else is renting intelligence and building nothing that lasts.

My grandfather did not rent his land in Alpine and Presidio. He worked it. He watched it. He built something that outlasted him.

That is the move.