The AI Implementation Report is a fortnightly newsletter for leaders navigating AI strategy. Each issue highlights the research, data and trends that matter most for implementation, not the hype, not the doom, just what is actually working and what is not. Curated by Mark Phelps and Drew Horton. A Mark Phelps LLC production.
Last issue we looked at what heavy AI use does to the people doing the work, and the fix that protects them. This issue steps back to the whole organization and asks a simpler question. Adoption numbers are climbing everywhere, so why is almost no one seeing a return.
THE ONE TO READ
If you only read one AI implementation article before your next leadership meeting, make it this one.
Adoption just crossed half the workforce. The number that predicts who actually gains is buried three paragraphs down, and isn’t adoption at all.
Organizational AI Adoption Jumps Six Points — Gallup (Jul 20, 2026) | FREE
For the first time in Gallup’s tracking, more than half of US workers use AI on the job. The Q2 2026 survey of 22,573 employed adults puts individual use at 52%, and organizational adoption jumped six points in a single quarter to 47%, the sharpest rise the series has recorded. If you have been waiting for AI adoption to become the norm, it’s official now.
So the adoption story is basically over, but the interesting number sits lower in the release. Daily use is 15% and frequent use, that is a few times a week or more, is 30%. Which means the other 70% of the workforce touches AI occasionally or not at all, and half of all workers use it a few times a year or less.
Here is the line that should reach your firm’s leadership team: Gallup finds that productivity gains scale with how broadly people apply the tool, not whether they have access to it. Among workers using AI for one or two kinds of task, about 45% report a productivity gain. Among those applying it across seven or more kinds of task, that climbs to 90%. Same tool, same license, double the return and the only variable is breadth of use.
That reframes your whole implementation problem. Access is solved, most organizations bought the licenses and can show a rising adoption chart. What almost none of them have is scope, people using AI across enough of their actual work to cross into real return. The gap between the firms winning with AI and everyone else is not who has it, It’s who uses it deeply, and that turns out to be a small band.
This usage span does not happen on its own. It comes from redesigned training, protected time to experiment on real work and managers who can tell people which tools are sanctioned and what data is safe. That is organizational work, that is planning and process design, the exact work most firms skipped on the way to a better adoption number to justify the investment.
So, the next time someone shows you an impressive adoption figure, ask the harder question underneath it. Not how many people have access, but how many use it across enough of their week to be in the 90%? That is the number that separates the firms leveraging the real value of AI and the organizations pulling ahead, from the ones that just bought seats and hoped for the best.
LEADERSHIP LOAD OUT: BRING THIS TO THE MEETING
Stop reporting AI adoption as a single percentage. Ask your functions to report breadth as well; how many distinct kinds of task each team actually applies AI to. A team at 60% adoption and one task type is not ahead of a team at 30% adoption and six. One of those is buying unused seats, the other is building capability, and only one of them shows up in the return.
ALSO ON OUR RADAR
Adoption is the easy half. Everything worth having sits on the far side of it, in the part almost no one has.
When Everyone Uses AI, Companies Risk Losing Critical Skills — Sagar Goel, David Martin and Charikleia Kaffe, BCG (Jun 17, 2026) | FREE
The other edge of universal adoption. BCG surveyed C-suite leaders and found about half already watching critical skills erode as AI absorbs the work that used to build them, and more than 60% expecting it to be a material threat inside three to five years. They call it distributed de-skilling, and their point is the one this whole newsletter keeps making: it is a design problem, not a tool problem. Breadth of use without deliberate design does not just fail to help, it quietly costs you the judgment you will need later.
AI Adoption Just Jumped Again. Are We Actually Using It Well? — Lead With AI (Jul 22, 2026) | FREE
A clean, short read on the same Gallup data if you want the argument without the full release. It lands on the capability overhang: the gap between having AI and using it well and why breadth rather than access is the number that moves productivity. **Useful to forward to the person in your firm who keeps citing the adoption percentage.
Rethinking Operating Models for Humans With Agents — David Mallon and colleagues, Deloitte Insights (Apr 2, 2026) | FREE
A signal of where this arc is heading. Deloitte finds 84% of companies have not redesigned jobs to fit AI even as they raise automation targets, and argues the fix is to make human judgment a design requirement rather than leaving people as the default catch-all for approvals and blame. That is the next question after breadth: once people are using AI deeply, who decides when it runs. We will pull at that thread in the issues ahead.
PARTING THOUGHT
Adoption is a number you can hit by buying licenses. Depth is a number you have to earn, one redesigned workflow and one protected hour at a time. The first one is easy, which is why almost everyone has it. The second one is hard, which is why almost nobody does, and why the few who do are pulling away from the rest.
Know someone navigating AI implementation? Forward this to them.
Hit reply and tell me what you think. We read every response.
And if there is something you think we should cover, send it over. We plan these issues as one connected argument, but the best suggestions come from readers, and we will give you props when yours lands.
Published for AI implementation leaders in the US and UK.
Mark Phelps is a product and design leader with more than 20 years of Fortune 100 experience at GE, Fidelity, TIAA, Bank of America and Marsh McLennan. As the founder of Mark Phelps LLC, he serves as a Fractional VP of Product Design with a focus on AI-driven environments.
For more than three years he has worked as a fractional product design leader, first at Rockefeller Capital Management, where he was engaged to provide critical process improvements, and then on an AI-powered healthcare software product, brought in for urgent user interface deliverables. Both engagements were extended well beyond their original scope. He is a named inventor on a pending design patent for the product’s interface.
He holds a Stanford AI Certificate earned with Distinction and an MBA in Digital Strategy from Suffolk University (Beta Gamma Sigma). He is a United States Marine Corps veteran.
20 Portsmouth Ave., Suite One #161, Stratham, NH 03885