AI in Fitness: Survey of 90 Certified Professionals | ISSA
ISSA Industry Research · AI in Fitness

AI in Fitness: What 90 Certified Professionals Told Us

The Fluency Dividend: the fitness professionals who use AI the most are the ones most confident that certification is becoming more valuable. The ones who have never tried it are the worried group.

Summary for editorsMethodology & limitsSurvey of 90 certified professionals · October 2025
Summary for Editors

What the topline concealed

In October 2025, ISSA surveyed 90 of its students and certified graduates about artificial intelligence. Topline results were published in January 2026. This paper reports what the topline concealed.

The headline finding

Among the 47 professionals using AI weekly or daily, 79% say AI will increase the value of holding a personal training certification over the next five years. Among the 17 professionals who never use AI, 24% say the same. The gap is 55 points, and it runs opposite to most commentary on this subject.

The finding that complicates it

Eighty-nine percent of frequent users report improved efficiency. Only 17% of all respondents report clear improvements in client outcomes. AI is returning time to fitness professionals. It has not yet been shown to make their clients healthier.

The finding nobody has acted on

Eight in ten respondents use AI in their own practice. Only 27% ever recommend it to a client, and 34% steer clients away from it. No certifying body has published guidance on where these tools belong in a client relationship.

Sample and limits

Ninety respondents, fielded October 30, 2025. A survey of one certifying body’s community, not a probability sample of the fitness industry. No margin of error is calculated. Skews are stated in the methodology section.

Interviews and dataset available. Contact Tami Nealy, tnealy@issaonline.com.

Why We Went Back

Averages hid the most interesting thing

ISSA published the topline results of this survey in January 2026, in an article by Tami Nealy, VP of Communications and Reputation. That piece reported what the whole sample said: 64% expect AI to increase the value of certification, 71% report improved efficiency, 64% say clients have never raised AI with them.

Those are the right numbers to publish first. They are also averages, and averages hide the most interesting thing in this dataset.

When you split the same respondents by how often they actually use AI, the 64% breaks into two populations that barely resemble each other. That split is the subject of this paper.

Finding One

A 55-point gap between users and non-users

Every respondent was asked how AI will affect the value of being a certified personal trainer over the next five years. Every respondent was also asked how often they use it. Crossing the two produces the clearest result in the survey.

How often they use AI Say certification will become more valuable
Weekly or daily (n=47) 79%
Monthly or rarely (n=25) 68%
Never (n=17) 24%

Subgroup bases sum to 89; one respondent did not provide a usage frequency. Weighted against the stated bases, these subgroup results reconcile to the 64% full-sample figure.

Across all 90 respondents, 64% expect certification to become more valuable and 13% expect it to become less valuable. Of the five people who said AI will greatly decrease that value, four had never used AI at all.

Sentiment tracks the same line. Forty-eight percent of daily users describe themselves as very optimistic about AI in fitness. Forty-one percent of people who never use it describe themselves as concerned that AI could reduce the role of trainers.

Two readings are available. Optimists may simply be more willing to try new tools. Or using the tools may be what teaches people their limits.

The written answers favor the second reading. Frequent users were the most specific about where AI fails, and the least worried about what that failure means for their careers. They described checking its output, correcting it, and treating it as a draft rather than an answer.

“AI can only be suggestions. The human has to use some brain power to discern whether what AI generates is useful in each situation.”
ISSA-certified professional, less than one year, gym or studio

That is not the answer of someone who has never opened the tool. It is the answer of someone who uses it and has learned to check it.

If the second reading is right, then anxiety about AI in this profession is a knowledge gap rather than a forecast. Knowledge gaps close with instruction.

Finding Two

Adoption runs backwards

The people you would expect to be most cautious are, broadly, the most fluent.

Years in the industry Use AI weekly or daily
Less than one year 68%
1 to 3 years 50%
4 to 7 years 50%
8 to 15 years 33%
15 years or more 45%

The pattern is directional rather than strictly linear: adoption falls across the first four tenure bands and then rises again among the longest-serving group. Fifty-nine percent of the sample has been in the industry less than three years, so the longer-tenure bands rest on small bases and should be read as indicative.

Delivery model matters even more than tenure. Coaches working fully online use AI weekly or daily at 83%. Coaches working fully in person do so at 38%.

This has a practical consequence for anyone in the education business. The people with the least professional experience are using these tools the most, with no guidance from any certifying body, employer or professional association. They are teaching themselves, and they are doing it during the period when their professional habits are forming.

That is an education gap, and closing it is straightforward work.

Finding Three

Six things professionals say cannot be replicated

We asked an open question with no options to choose from: what aspects of your work do you believe cannot be replaced or replicated by AI? Seventy-five people wrote an answer in their own words. Coded by theme, they cluster into six competencies.

  • Connection and trust (48% of answers). The most common answer by a wide margin. Building rapport, earning trust, and knowing a client well enough that the relationship itself is the thing that works.
  • Hands-on coaching (24%). Watching a movement, correcting form in the moment, and giving physical feedback a screen cannot deliver.
  • Empathy and reading the person (20%). Noticing that a client is not present today, and changing the session because of it.
  • Accountability (17%). A plan is easy to produce and hard to follow. Following it is where a coach earns the fee.
  • Real-time judgment (15%). Deciding when a client is ready to progress, when to back off, how to work around an injury, and how to adapt a program to a body that does not match the template.
  • Presence and energy (7%). The least measurable answer and one of the most consistently expressed. Several professionals described the effect of their own presence on a room as part of what clients pay for.
“AI may be able to generate workouts, but it does not have the ability to gain and build rapport like people do. It cannot distinguish microexpressions during a movement, nor provide any sort of tactile feedback. AI cannot read stress. We know when a client is just not there and may need to talk through some stresses and change the workout to a mobility session or a gentle release to help them mentally.”
ISSA-certified professional, 15+ years, independent
“What is extremely hard is the adherence to the plan, not building the plan itself. If the customer needs a plan, maybe AI would be helpful. But if they are looking for results, they really need a coach.”
ISSA-certified professional, 1-3 years, independent

Read together, these six describe a job rather than a list of things machines happen to be bad at this year. Each one requires a second person in the room, paying attention, and willing to be responsible for what happens next. None of them is a knowledge problem that a better model solves.

They are also all teachable, which is the point. Assessment, cueing, behavior change and program adaptation are curriculum. The competencies that survive this technology are the competencies a certification exists to build.

Finding Four

Time, not results

Among professionals using AI weekly or daily, 89% report improved efficiency and 55% report significant improvement. Time savings are real, they arrive quickly, and they explain why adoption keeps climbing.

Client results look different. Across all respondents, 17% report clear improvements in client outcomes such as goal achievement or consistency. Another 12% report measurable improvements in business performance. The rest report minor improvements, no change, or say it is too early to tell.

Where the time goes is easy to see.

What professionals use AI for Share
Program design and workout planning 54%
Marketing and social media content 41%
Nutrition guidance 36%
Continuing education and learning 29%
Client progress tracking and data analysis 26%
Administrative tasks 26%

Respondents could select more than one purpose.

Sixty-eight percent use ChatGPT or a similar chat assistant, and no other category comes close. AI-enabled fitness and nutrition apps sit at 29%, marketing and content tools at 21%, and 19% report using nothing yet.

So the current state of AI in coaching is this. It drafts, it summarizes, it saves hours, and it has not yet been shown to improve anyone’s health. What a professional does with the hours it returns is still the whole job.

Publishing this alongside the efficiency number matters. A paper that reports only the flattering figure invites a journalist to find the other one.

Finding Five

What professionals use, and what they recommend

Eight in ten respondents use AI in their own practice at some frequency. Only 27% ever recommend an AI tool to a client. Thirty-four percent say they would rather clients focus on human-led guidance.

Do you recommend AI tools to clients? Share
No, I prefer clients focus on human-led guidance 34%
Not yet, but open to it 31%
Occasionally, when relevant to a specific goal 16%
Yes, frequently 11%
Rarely, mentioned but not encouraged 6%

The profession has arrived at a working position that nobody wrote down: these tools are useful for professionals and risky for clients. That position may well be correct. It has never been examined, and no certifying body has published guidance on it.

Add the silence on the other side of the relationship, where 64% of respondents say clients have never raised AI with them at all, and the picture is a profession that uses these tools daily, holds strong views about them, and discusses none of it with the people paying for the service. That silence is examined in the companion paper, The Silent Fitness Client.

Implications

What follows from this

  • Fluency belongs in continuing education. The 55-point gap is a training gap. It does not close with reassurance.
  • The client conversation has to start. Professionals should be asking clients what advice they are already following, because most are not asking and most clients are not volunteering it.
  • The profession needs a standard. Not a policy imposed from above, but one built from what practitioners already do, published openly, and revised in public.
  • Research should be continuous, not occasional. This survey is one wave, with 90 responses and clear limits. The value of the finding is in watching it move.
Methodology

Instrument, population and limits

Instrument. An 18-question online survey covering certification held, tenure, practice type, AI usage frequency, tools used, purposes, reported effects, and open-ended questions on the limits of AI in coaching.

Population. ISSA students and certified graduates. Participation was voluntary and responses were anonymous.

Fielded. October 30, 2025. Responses. 90.

Composition. 86% hold the ISSA Certified Personal Trainer credential; 51% also hold Nutritionist and 39% Strength and Conditioning. 59% have been in the industry less than three years. 67% deliver coaching mostly or entirely in person. 44% are independent trainers or coaches and 20% work for a gym or studio.

Limits. This is a self-selected sample from a single certifying body’s community. It is not a probability sample of the fitness industry and no margin of error is calculated. The tenure and delivery-model skews are stated wherever they affect a finding. Open-ended answers were coded into themes by a single reviewer, and answers containing more than one theme were counted in each. Percentages are calculated on all 90 responses unless a subgroup size is stated. Quotations are reproduced as written, with spelling corrected and nothing else changed.

Prior publication. Topline results were published by ISSA in January 2026 as “The Human Advantage: How AI Is Reshaping, Not Replacing, Personal Training.” This paper reports the subgroup analysis that piece did not cover.

Next wave. ISSA will field a second wave with a larger sample and a stated margin of error, with the intention of publishing an annual index of AI adoption among certified fitness professionals.

Cite this paper

Nealy, T. The Fluency Dividend: Why the Fitness Professionals Using AI Most Are the Least Worried About It. ISSA Industry Research, 2026. issaonline.com/pages/ai-in-fitness-survey

Reviewed by Jenny Liebl, CPT. This paper may be quoted with attribution to ISSA. Interviews and the de-identified dataset are available on request.

Questions

About this research

Do fitness professionals think AI will replace them?

Mostly no, and the answer depends heavily on whether they use the tools. In ISSA's October 2025 survey of 90 certified and aspiring professionals, 79% of those using AI weekly or daily (n=47) said AI will increase the value of holding a personal training certification over the next five years. Among the 17 respondents who never use AI, 24% said the same. Across all 90 respondents the figure was 64%.

What do personal trainers use AI for?

Program design and workout planning leads at 54%, followed by marketing and social media content at 41%, nutrition guidance at 36%, continuing education at 29%, and client progress tracking and administrative tasks at 26% each. 68% use a general chat assistant such as ChatGPT, and no other tool category comes close.

Has AI improved client outcomes in fitness?

Not yet demonstrably. Among professionals using AI weekly or daily, 89% report improved efficiency and 55% report significant improvement. But across all respondents only 17% report clear improvements in client outcomes such as goal achievement or consistency. AI is returning time to fitness professionals; it has not yet been shown to make their clients healthier.

Do trainers recommend AI tools to their clients?

Rarely. Eight in ten respondents use AI in their own practice, but only 27% ever recommend an AI tool to a client, and 34% say they would rather clients focus on human-led guidance. No certifying body has published guidance on where these tools belong in a client relationship.

What is the sample size and margin of error?

Ninety responses, fielded October 30, 2025, among ISSA students and certified graduates. Participation was voluntary and responses anonymous. This is a survey of one certifying body's community, not a probability sample of the fitness industry, and no margin of error is calculated.

About ISSA

International Sports Sciences Association (ISSA) is the world’s most connected fitness and wellness ecosystem. Since 1988, ISSA has gone beyond traditional certification to build the industry’s career operating system, giving professionals the skills, confidence, and connections to build meaningful futures, while providing operators with the trusted, work-ready talent they need to grow. Its premier workforce platform, Career Connect, bridges a global community of over one million students across 176 countries with more than 23,000 global partner locations. ISSA aims to connect 100 million people to healthy living by 2030.

This paper may be quoted with attribution to ISSA. Media enquiries: Tami Nealy, VP Communications & Reputation, tnealy@issaonline.com.