AI and Sports Medicine: What Providers Need to Know

Patients are already asking ChatGPT and Claude about their bodies. Here's what AI and sports medicine providers need to know to stay ahead.

In the last seven months, OpenAI, Anthropic, and Perplexity each launched tools that let patients connect their health data directly to an AI assistant. It's easy to have missed why this matters for your practice.

In January, OpenAI launched ChatGPT Health, a dedicated space where users connect medical records, Apple Health, MyFitnessPal, and wearable data so ChatGPT can answer questions grounded in their own history. The January rollout landed as a limited pilot, and by most early accounts, a fairly quiet one. That changed on July 23, when OpenAI opened it to every U.S. adult on every plan, free included, one day after a Florida pastor sued the company over an alleged near-fatal medical suggestion ChatGPT had given him (TechCrunch). By that point, OpenAI said the tool was fielding roughly 300 million health and wellness questions a week, up from the 230 million it had reported back in January (TechCrunch).

That timeline matters. This isn't a clean, uninterrupted growth story, it's a product that had a rocky start, drew a lawsuit, and got pushed to hundreds of millions of people anyway. Worth knowing, since your patients are the ones now weighing what to trust.

OpenAI isn't alone. Anthropic launched its own healthcare push the same week in January: connectors that let Claude pull from Apple Health and Android Health Connect, plus medical records from more than 50,000 provider organizations nationwide through a partnership with HealthEx (Fortune). That access is currently limited to Claude Pro and Max subscribers, not the free tier. Fortune's own reporting noted Anthropic has been less focused on the general consumer market than OpenAI, whose free tier alone reaches over 800 million weekly users. In March, Perplexity introduced a similar suite of health connectors, Apple Health, Fitbit, Ultrahuman, Withings, and electronic health records from over 1.7 million providers, also starting with Pro and Max subscribers (Perplexity).

Three of the most closely watched AI companies in the world independently concluded, within months of each other, that connecting personal health and fitness data to a conversational AI is where consumers are headed next. Their reach isn't identical, OpenAI's move touches its entire free user base while Anthropic's and Perplexity's currently reach paying subscribers only, but the direction is the same across all three, and that's worth paying attention to.

What this actually means for physical therapy and sports medicine providers

None of this means your PT clinic, sports med practice, or performance program is about to be replaced by AI. Reviewers and industry analysts have been consistent on this point: ChatGPT Health is explicit that it's "designed to support, not replace, medical care" and isn't built for diagnosis or treatment (OpenAI), and one healthcare-strategy analysis noted the feature isn't technically groundbreaking, it's mostly a more convenient way to do things people were already trying to do with these tools (HMA).

But convenience is exactly the point. Picture a client who used to arrive at your clinic with vague complaints and a rough sense of their training history. Increasingly, that same client shows up having already asked ChatGPT to compare this month's sleep data against last month's, cross-reference it with a nagging knee issue, and suggest what to bring up at their appointment. A WestBridge Capital investor put it bluntly to MedCity News: broad, chat-based wellness and nutrition products are heading toward commoditization now that Claude and ChatGPT have that kind of reach and habitual daily use, while businesses built around specialized clinical expertise and human-clinician relationships are the ones positioned to hold their ground (MedCity News).

That's the real shift. It's not "AI replaces physical therapy." It's "the baseline for what a well-informed, data-connected patient expects from any health interaction just moved up," and providers whose value proposition was "we have information you don't" now need a new answer.

The insurance-backed MSK players were already circling

This pressure isn't arriving in a vacuum. For the past several years, digital musculoskeletal (MSK) care companies have been steadily pulling volume away from traditional and independent PT clinics by getting embedded directly into employer health plans.

Hinge Health and Sword Health are among the biggest names in this category, and they compete hard on the exact things AI is now amplifying: data-driven, always-available care. Hinge pairs an app-based exercise program with health coaches and reported 25 million contracted lives as of its most recent quarterly results, up from 20 million a year earlier (Hinge Health Q4 2025 results). Sword leans on 100% licensed Doctor-of-PT-delivered care paired with an FDA-listed motion-sensor device for real-time biofeedback on every exercise (Sword Health). A peer-reviewed study in *JMIR Rehabilitation and Assistive Technologies* found participants' pain scores dropped 73% by the 12-week mark, compared to a group of registered users still waiting on their benefits to activate (Wang et al., JMIR). Worth noting: it's not independent research, most of the study's authors are Hinge Health employees with equity in the company, though the study did clear external peer review.

These programs are attractive to employers and insurers because they're free or low-cost at the point of care, scale without hiring more clinicians, and produce clean outcomes data. For the patient, the pitch is "your insurance already covers this, and it's easier than driving to a clinic." That's a hard combination to compete with on convenience or price alone, and now that same patient population is also getting free, always-on AI analysis of their wearable and health data through ChatGPT, Claude, or Perplexity. The two trends compound: insurance is subsidizing the tech-enabled path, and consumer AI is making people feel more capable of self-directing their own care before they ever set foot in a clinic.

Why this cuts both ways for independent and cash-pay providers

Here's the tension worth sitting with. Independent and cash-pay physical therapy practices already differentiate from high-volume, insurance-billed clinics on exactly the dimensions that matter most right now: one-on-one time with a licensed clinician instead of rotating through aides, 45- to 60-minute sessions instead of 15- to 20-minute ones, and care built around an individual's goals rather than a standardized protocol dictated by reimbursement codes (Physical Therapy Doc, Pabau). It's a genuinely strong position against a one-size-fits-all app-based program.

But it's also a position under new pressure from two directions at once:

• From above: Insurance-backed MSK vendors are normalizing the idea that "free, app-based, data-rich care" is the default expectation, which raises the bar for what any provider, including a premium cash-pay one, needs to visibly deliver to justify a higher price point.
• From the side: Consumer AI tools are giving clients a taste of personalized data analysis for free, which chips away at one of the things clients used to pay a provider specifically to interpret for them.

The providers who lose ground here are the ones whose value proposition was primarily "we have your data and the tools to read it," because that layer is exactly what's getting commoditized. The providers who gain ground are the ones who can show, concretely, that the relationship, judgment, and program design around that data is worth paying for.

How physical therapy and sports medicine providers can adapt

This is the part worth taking seriously, because "differentiate on the human relationship" is true but not actionable on its own. A few concrete directions:

1. Make the AI-informed patient part of your intake, not a threat to it.

Patients are going to walk in having already asked ChatGPT or Claude about their symptoms, training load, or wearable trends. Build an intake process that explicitly invites that context in rather than starting from zero, ask what they've already looked into, what the AI told them, and use that as a starting point for a real clinical conversation. This reframes the provider's role from "information gatekeeper" to "the person who can actually validate, correct, and act on what you've already learned," which is a stronger and more honest position anyway.

2. Build (or repackage) programs around outcomes the consumer tools can't produce.

General-purpose AI can summarize a lab result or flag a sleep trend. It isn't built to run a return-to-play protocol, manage in-season load across a roster, or make a clinical judgment call about when an athlete is truly ready to compete again. Programs that are explicitly structured around those higher-order, judgment-intensive services, not just "we'll look at your data with you", are much harder to commoditize.

3. Turn transparency into a selling point.

If patients are going to bring AI-generated summaries and self-diagnoses into the room regardless, providers who proactively show their own data-driven reasoning, here's what we're tracking, here's why, here's how it's different from what generic AI would tell you, build more trust than providers who treat outside AI use as a nuisance to route around.

4. Compete on integration, not avoidance.

Employer-backed MSK apps win partly because they're frictionless. Independent and cash-pay practices can match that friction reduction, better scheduling, clearer communication, data that follows the patient between the field, the gym, and the clinic, without giving up the individualized, high-touch care that's the actual differentiator. The goal isn't to out-app the apps; it's to make sure convenience isn't the reason someone chooses a lower-quality, standardized program over a better one.

5. Know where the insurance-backed model has real downsides, and say so.

Programs like Hinge and Sword are compelling to employers precisely because they scale without much clinician involvement, which for many patients means less individualized attention, more one-size-fits-all protocol, and less flexibility if a case doesn't fit the standard pattern. It's a legitimate trade-off, and providers who can clearly and fairly articulate it, not as a knock on those programs, but as an honest comparison, give price-sensitive or insurance-eligible prospects a real reason to choose a more tailored path anyway.

The bigger picture

AI connecting to personal health data isn't a distant trend anymore, it's live, mainstream, and moving fast enough that three separate AI companies built competing versions of it within the same seven-month window. For physical therapy and sports medicine providers, the smartest response is neither to ignore it nor to compete with it head-on. Get clear about what a clinician actually offers that AI, even with access to your Apple Health data, doesn't, and build (or rebuild) programs that make that value obvious from the first conversation.

At Rivalists, this is the exact problem we spend our time on, helping sports medicine and physical therapy providers structure their data, their programs, and their client experience so the human expertise stays the most valuable part of the relationship, no matter how smart the tools around it get. If you're new to the field or want the fuller picture of how these roles fit together, our guides to what sports medicine covers and who's actually on a sports medicine team are a good place to start.