In January, Playlist and EGYM announced a $7.5 billion merger backed by $785 million in fresh equity, with new investment from Affinity Partners, Vista Equity, Temasek, and L Catterton. The combined entity (Mindbody, ClassPass, Booker, and EGYM under one roof) generated over $800 million in net revenue in 2025. A big chunk of that new funding is going straight into artificial intelligence.
It's not really a fitness deal. It's a data and AI deal that happens to involve gyms.
And it's not happening in isolation. Over the past few weeks, the signals have been coming thick and fast. Virgin Active has hired a Global AI Director - a deliberate, senior appointment to drive AI strategy across the business. Louise John, their European Head of IT Service Delivery, put it plainly: they could "put AI into everything and not get the value back." So they've hired someone whose entire job is making sure they do.
Life Time, fresh off nearly $3 billion in FY2025 revenue, launched L.AI.C (pronounced "lay-see" apparently), an AI wellness companion built on Azure AI Foundry, trained on 30 years of Life Time's health and fitness content. It's free. Not just for members. For everyone (provided you sign up). That's a data acquisition strategy dressed up as a health tool.
PureGym, meanwhile, has gone 100% digital on their contact centre and is rolling out AI to spot members entering clubs without subscriptions, optimise gym layouts, manage equipment quantities, and identify when kit is out of order.
Basic-Fit has deployed AI "Fitbuddies": five distinct AI personas designed around behavioural data from 4.4 million members, targeting the 4-6 week window when most new members drop out.
Technogym's AI Ecosystem now sits on 38 million connected users. Technogym claims operators using the platform see +30% in acquisition, +20% in retention, and +100% in secondary spend.
These aren't just experiments. These are strategic bets by operators and suppliers who've decided AI is the next competitive battleground.
But Members Don't Want AI. Right?
Here's where it gets interesting.
The Les Mills 2026 Global Fitness Report surveyed 10,000 consumers across five continents. Only 10% said they'd prefer an AI coach. 52% actively prefer a human trainer. And the generational breakdown is the real surprise: Gen Z (11%) are more resistant to AI coaching than over-55s (13%). The supposedly tech-native generation wants a person, not an algorithm, when they're working up a sweat.
Equinox picked up on this: their 2026 brand campaign is built around challenging AI ("Question Everything But Yourself"), explicitly positioning the human body and human expertise as the antidote to a synthetic world.
So the industry's biggest consumer survey says members don't want AI. And one of the most influential premium brands is actively campaigning against it. Does that blow a hole in the AI argument?
No. It sharpens it.
The AI That Matters Isn't the AI Members See
The operators making serious AI investments aren't building robot personal trainers. They're building operational intelligence.
PureGym isn't using AI to coach members through a squat. They're using it to figure out that three treadmills on the second floor haven't been used much this month and the cable machines on the ground floor have a 10-minute wait at 6pm. That's a layout problem with data behind it — something you can make real decisions against and actually measure.
Life Time isn't replacing its trainers with L.AI.C. It's capturing health and wellness interaction data from millions of users (members and non-members) that feeds back into how they design clubs, structure programmes, and predict what services to invest in next.
Virgin Active didn't hire a Global AI Director to build a chatbot. They hired one because, as Louise John acknowledged, AI without focus is just cost without return. The job is to work out which operational problems AI can actually solve and make sure the investment pays back.
This is the distinction most operators are missing when they focus on AI purely for member engagement. The Les Mills data is telling them something important — just not what they think. It's not saying "AI doesn't matter." It's saying "AI-as-a-personal-trainer doesn't matter." The AI that matters is behind the scenes: churn prediction, space optimisation, dynamic pricing, equipment lifecycle management, energy cost modelling, member journey mapping. The boring stuff. The stuff that makes a 50-site chain significantly more efficient while their competitors are still running on spreadsheets and gut feel.
The Data Debt Problem
Here's the hard bit: you can't deploy operational AI on data you haven't captured.
Every gym that's still running on manual check-ins, paper-based maintenance logs, and monthly spreadsheet reviews is accumulating data debt. They're not just behind on technology — they're behind on the raw material that makes the technology work. When they eventually decide to adopt AI (and they will, because their competitors already are) they'll discover they've got nothing to train it on.
The operators investing now aren't just getting smarter today. They're building a compounding advantage. Every week of member behaviour data, every equipment usage pattern, every energy consumption reading, every class attendance record — it's all fuel for models that get better over time. The gap between operators who started collecting this data two years ago and those who start next year isn't two years. It's exponential.
Who Gets a Pass?
If you're running a single gym where the draw is you — your personality, your coaching, your community — AI probably isn't going to change your business model. Your members come for the relationship. They stay because of the relationship. An AI can't replicate that, and it shouldn't try.
But if you're running anything bigger than that (a multi-site chain, a franchise, a leisure trust with a dozen facilities and a council contract to maintain) the question isn't whether AI will matter. It's whether you'll have the data infrastructure and organisational muscle in place when it does.
The Playlist-EGYM deal puts $785 million behind the bet that you will. Or more precisely, that enough operators will to make the platform worth building. The ones who aren't ready when that platform arrives will find themselves choosing between adopting someone else's AI on someone else's terms, or competing without it entirely.
Neither option is great.
