AI news & training for business owners & operators. One email. One clear next step.

THIS WEEK IN AI

I took a few weeks off to rebuild this newsletter from the ground up. You are reading the new version. Shorter, sharper, and built around one idea per issue.

This week, Gallup put a number on something I have been watching for a while. Half of U.S. workers now use AI at work. But the results inside most companies are not matching the adoption curve. That is the story I want to walk through with you today.

What happened:

Gallup surveyed 23,717 U.S. employees in February. For the first time, half now say they use AI at work at least a few times a year, up from 21% in 2023. Daily and weekly usage hit an all-time high of 28%. But only about one in ten employees at companies that have adopted AI strongly agree that it has changed how work actually gets done.

Why it matters to your business:

The headline is adoption. The real story is the gap. 67% of leaders at AI-adopting companies use AI daily or weekly. Only 46% of individual contributors do. The people you pay to do the work are using AI less than the people you pay to direct it. And 27% of employees at AI-adopting companies say their workplace has been disrupted in a major way in the past year. Tools are spreading. Clarity is not. If you are leading a company with 20 or 200 employees, the question is not whether your team is using AI. It is whether anyone on your team knows what they are supposed to be using it for.

Source: Gallup Workplace report

What happened:

PwC surveyed 1,217 senior executives across 25 sectors for its 2026 AI Performance Study, released April 13. The top 20% of companies are capturing nearly three-quarters of AI's financial gains. The other 80% are sharing the remaining 26%. Leaders generate 7.2 times more AI-driven revenue and efficiency gains than the average competitor.

Why it matters to your business:

Read that again. This is not about how much AI a company uses. It is about what the company points AI at. PwC found the leaders are using AI to redesign workflows and drive growth. Everyone else is layering AI on top of existing processes and calling the pilot a win. If your AI strategy is "let everyone try ChatGPT and see what happens," you are in the 80%. The 20% built the foundations first.

Source: PwC 2026 AI Performance Study

What happened:

Anthropic released Claude Opus 4.7 on Thursday, its most capable generally available model. Gains over the previous version include stronger software engineering, better long-running task reliability, sharper instruction following, and higher-resolution vision. Pricing stayed flat at $5 per million input tokens and $25 per million output tokens. The release comes about two months after Opus 4.6 and puts Anthropic back ahead of OpenAI's GPT-5.4 and Google's Gemini 3.1 Pro on several benchmarks.

Why it matters to your business:

You do not need to care about the benchmarks. You need to care about the pace. The major AI companies are now shipping meaningful model upgrades every two months. If your team evaluated AI tools six months ago and decided "it is not ready yet," that decision is already stale. The gap between what these tools can do and what your business is using them for is getting wider every quarter. This is not a reason to panic or to buy more subscriptions. It is a reason to have a standing quarterly review of what your team is actually using and whether the version you picked last year is still the right one. Stop treating AI tool selection like a one-time decision. Treat it like the recurring review you already do for insurance, software, and vendors.

Source: Anthropic announcement

I posted about the Gallup number on LinkedIn this week. The response told me something. A lot of people recognized themselves in the gap between "we use AI" and "AI changed our results."

Here is what I want to say to the people in that gap.

The story the headlines are telling is that AI adoption is finally here. Half of workers are using it. We made it. And if you read that story and feel behind, you are going to do what most leaders do when they feel behind. You are going to buy more tools, push your team to use them, and wait for the results to show up.

The results will not show up. Not from that.

Look at what PwC actually measured. They did not ask which companies had the most AI tools. They asked which companies were getting the most revenue and efficiency gains from AI. And the answer was brutal. 20% of companies are capturing 74% of the value. The other 80% are running pilots, producing reports, and watching the gap widen.

The leaders are not winning because they bought more. They are winning because they pointed AI at work that was already organized enough for AI to help. That is the part nobody wants to hear, so it is the part I am going to say plainly.

AI amplifies the workflow you already have. If your estimate process is messy, AI gives you messy estimates faster. If your handoffs are unclear, AI generates more unclear handoffs. If nobody owns the output, AI produces more unowned output. You cannot automate your way out of a process that was never mapped.

I worked with a contractor last year whose estimate process took three hours. We sat down and walked through it. Not the AI part. The whole process. Who touches it, what they need, what they hand off, what gets redone. After we cleaned that up, we added AI where it actually fit. The process now takes 20 minutes. That is not an AI success story. That is a workflow story with AI in it.

The reason this matters right now is that the Gallup data and the PwC data are telling you the same thing from two different angles. Adoption is up. Results are stuck. And the longer that gap stays open, the more the market separates into the 20% pulling ahead and the 80% paying for subscriptions that are not changing anything.

If you are in the 80% right now, that is normal. Most businesses are. The fix is not more tools. It is looking at one process in your business and asking an honest question before you add anything to it. Where does this actually break down? Who owns it when it does? What would we need to clean up before AI could help here?

Fix the work first. Then add AI.

That is it.

Most leaders I talk to assume their team is either all-in on AI or not using it at all. The truth is usually messier. Someone in accounting is using ChatGPT to write emails. Someone in operations has never opened it. Someone in sales has a personal subscription they pay for themselves because they do not want to ask permission. You cannot build an AI strategy on top of assumptions about what your team is doing.

Listen to what they say. Write it down. You are not looking for a right answer. You are mapping the actual state of AI use inside your business. That map is the starting line for every decision that comes next.

I rebuilt this newsletter from scratch. If it was useful, forward it to someone on your team who is trying to figure out AI for their business. And if you want shorter takes like this during the week, follow me on LinkedIn: CLICK HERE!

Thanks for reading,

Andrew Keener

Operations & AI Strategist

Keen Alliance Consulting