FIELD NOTES | Ontario Utilities | September 2026
The 70 Percent Nobody Funded
What Ontario utility leaders are learning about where the return on AI actually lives. In partnership with MEARIE.
Every utility leader looking at AI is holding two questions at once, in the form of two separate and related conversations.
The first is a money question. We have spent on licences, and we have run some training. Where is the return?
The second is a people question. Our crews, customer service team and engineers are watching this closely. Some are curious, others are quiet. Almost everyone is nervous about AI. How is this impacting our culture?
These are not two separate conversations. They are the same problem, and the evidence proves it. The return on your AI investment sits inside the people question, which is why almost everyone rolling out an AI plan has yet to see real ROI. Yet.
The money question
Every AI investment moves through three stages. First, you buy the tools and secure enough licences for your team. Then you adopt them, meaning a rollout and some sort of training. Then comes the stage that decides whether any of it pays off. You adapt, changing how your people actually work.
- Enterprise spending on generative artificial intelligence reached $37 billion in 2025, up from $11.5 billion in 2024. (Menlo Ventures, 2025: The State of Generative AI in the Enterprise, December 2025)
- 74 percent of companies have yet to show tangible value from their use of artificial intelligence. Only 26 percent have built the capabilities to move past pilots. (Boston Consulting Group, AI Adoption in 2024, October 2024)
- Only about 5 percent of generative artificial intelligence pilots are delivering a measurable return. The Massachusetts Institute of Technology calls the cause a learning gap, not a technology gap. (MIT NANDA, The GenAI Divide: State of AI in Business 2025, August 2025)
- 56 percent of chief executives say artificial intelligence has delivered neither higher revenue nor lower cost in the past 12 months. (PwC 2026 Global CEO Survey, January 2026)
- The organizations winning with artificial intelligence put 10 percent of their effort into algorithms, 20 percent into technology and data, and 70 percent into people and processes. (Boston Consulting Group, AI Adoption in 2024, October 2024)
That last one hits us the hardest. If 70 percent of the AI value lives in people and processes, and your budget is almost entirely licences and a vendor rollout, the gap in return is not a mystery. It is a line item not properly funded.
Tactics with no strategy get you no return. Strategy with no tactics gets you no return. Your people are the bridge between them.
The culture question
AI is a cultural conversation before it is a technological one.
Hesitation about AI is almost never about the technology. It is about whether people believe they still matter once it arrives. When a team feels overlooked, or something is being forced upon them, AI is almost always viewed as a threat, meaning adoption stalls in ways that never show up in a usage report. When a team feels seen and trusted, they use, test and improve the same tools, and they tell you what is working and what is not.
This is the hard work and leadership behind the 70 percent. Culture is what converts a licence into a changed way of working. Training helps, and it is not enough on its own.
Ontario utilities are well positioned to adapt with AI. Reliability, safeguarding critical infrastructure and public trust are already how the sector operates. Those instincts do not disappear with AI. They are the judgment AI cannot supply, and utilities that embrace AI as a cultural conversation will get more out of the same tools than most enterprises will.
The practice: See, Think, Co-Create
The +ai Mindset is a practice, not a framework to memorize. It is built on three abilities your teams already have and need to reimagine.
See
AI sees patterns. Your people bring the meaning.
AI finds connections in data with remarkable speed and scale, and it does not understand what it is looking at. It does not know why reliability matters. It cannot feel the weight of public trust or the pressure of an unplanned outage.
Your grid operators see system health, not just voltage readings. Your compliance team sees regulatory intent, not just rule changes. Your customer service team sees worried neighbours, not just account numbers. AI sees probability. That is not a limitation, it is a healthy division of labour, and the future of work.
The early AI wins we are seeing are rarely moonshot projects. They are the monthly vegetation report that takes two days, repetitive billing enquiries, and the compliance spreadsheet three people maintain separately.
Think
AI does not think. Your people orchestrate it.
When utilities hit roadblocks with AI, it is rarely the technology. It is the compliance team that will not let AI draft a filing in case it is wrong. It is the veteran engineer who does not trust AI for load calculations. It is the customer service manager who has done the math on what this means for her team.
Behind each of these scenarios is a person protecting something important. A +ai Mindset knows the job is not to argue them out of it. It is to understand the expertise they are defending, address the risk they are avoiding, and show them where AI makes their judgment go further. Your team did not become experts by knowing all the answers. They became experts by asking better questions.
Co-Create
Not automation. Collaboration.
The most common mistake is treating AI like previous automation projects. Here is a task, do it for us. That approach hands over the work, keeps none of the judgment, and is why so many pilots produce faster versions of mediocre output.
The utilities getting real value do the opposite. The regulatory team drafts a first version with AI and edits it with human expertise. The communications team builds message templates and personalizes them. The operations team generates several planning scenarios and applies judgment to choose. Reports get better, not just faster. Customer communications get more consistent and more human.
Underneath all three sits the question that never changes. Who are we becoming with AI? While the tools will keep changing, a +ai Mindset gives your team the skills to change and grow with them.
What this looks like in an Ontario utility
The utilities making this stick are the ones treating iteration as a chance to get better. This could look like:
- An AI champion named in each department, chosen for curiosity rather than technical skill
- Permission to experiment on low-risk work, with clear boundaries
- Regular sessions where teams share the AI interactions that worked and the ones that missed
- Leadership that celebrates the experiments that fail instructively
- A way to measure confidence before and after, so the change is visible to the people funding it
The Ontario utility case study, coming this winter
We are co-authoring a case study with MEARIE and an Ontario member utility, following a full training engagement from first conversation through delivery and into the months after, with team confidence measured at every stage. We are writing it in partnership with both.
We believe hard data helps leadership teams move through buy, adopt and adapt faster and with more confidence. Right now many organizations are making decisions on vendor claims and anecdotes. It deserves numbers from a peer utility.
If you would like it when it is ready, send us your name and email and we will make sure you have it.
What this could look like next for your team
90 minutes
The conversation your team is already having, out loud.
Three hours
A hands on session to explore what your tools do with real work in real time.
Full day
Align the whole team at once, in your space and on your tools.
Twelve weeks
Build new habits so your AI licences generate real ROI and efficiencies.
Everything above adapts to the tools you already run. We work in Copilot, ChatGPT, Claude and Gemini every day, so we come into your environment rather than asking your team to come into ours. Who are you becoming with AI? Let us find out together.
About the authors
Sean Yo
Sean created the +ai Mindset and leads its thinking and its writing. Three decades in technology, from big data platforms to launching PwC’s internal ventures lab to growing a college innovation centre from one person to more than eighty. He was early to artificial intelligence because he saw it as an equity and opportunity conversation rather than an automation one, and he still teaches it that way.
Curt Hammond
Curt is the founder of Four Simple Words and a long-time MEARIE trainer. He has spent more than 20 years helping organizations turn big ideas into something a whole team can use, across Fortune 500 companies, government and small businesses. When GPT-4 launched, his online naming business lost 80 percent of its revenue in six weeks. That is when he stopped using artificial intelligence and started building with it.
Together they work as Four Simple Words, a service leadership practice built on a single idea. People do their best work when they know that they and their ideas matter.
How these field notes were made
We wrote this with AI, using our own material and evidence bank. Every number above is sourced, linked and checked by hand, so the facts are indeed facts.
Then we did the thing we teach. We used AI to see patterns across more session notes and client conversations than we could hold in our heads at once. We used it to think, arguing with it about what actually mattered and cutting far more than we kept. And we co-created this, draft after draft, with our judgment sitting on every line.
That last part is the real work, and most people do not see or understand it. The early drafts were not good. Getting here took real hours from both of us, and it is better than either of us would have written alone.
Our workflow has become tight enough that we cannot always tell you where Sean ends, Curt begins and Claude helps. We can tell you exactly who is accountable for every word. That is us.
