Show Notes:
In this episode, we close out the two-part series on AI in ACL rehab with the territory that matters most. We open with a use case we did not cover in part one, using AI as an audit tool to compare your actual care against what the research says, and why the way you prompt it determines whether you get honest information or just the answer you were already looking for. We then finish tier two with the pattern of athletes who build a program from AI and troubleshoot it with AI, no human ever in the loop, and move into tier three, the hard lines where AI should never be the decision maker. We break down exactly what AI cannot know when it builds your program, your graft type, your quad deficit, your swelling pattern, your movement quality, and your psychological readiness, and why the return to sport decision requires a human who can test you, watch you move, and take responsibility for the call. We get specific about how we use AI at The ACL Athlete, where it earns its place in our workflow and where it does not touch the process at all. We close with the three question framework you can use every single time you reach for AI to answer something about your knee, your program, or your recovery.
What is up team, and welcome back to another episode on the ACL Athlete Podcast. Today is part two of this series on using AI, artificial intelligence, in ACL rehab. If you have not listened to part one, go back and start there. This builds directly on it, and I really encourage you guys to listen because it is going to be very important as we talk through this episode.
Especially in general, this series is so important because AI is going to be used so much more as time goes on. It is going to be at the forefront of how people get answers to their questions and how they search for things. Google itself automatically has an AI-based summary overview that it creates. So even if you Google something, you are getting things from AI.
Yes, you can go to specific links, but when people are searching these days, AI will be involved in the process, and it will become second nature in terms of accessing information.
With this podcast episode series, this is based on current 2026 realities. We will see how this changes as time goes on because this stuff is rapidly changing so, so fast.
This is something I want to be helpful for you guys as you listen, because it is a bit of a challenging space to navigate. As I shared in part one, AI is pulling from the internet. It is pulling from the information it has been fed. It is not vetting whether something is legit or not legit. It is just information it can pull and aggregate together based on a large language model. That is what AI is.
For a quick recap of part one, we covered where AI is genuinely useful: understanding your diagnosis, preparing for surgical consultations, navigating the healthcare system, and research support for clinicians. We introduced the three-tier framework and got into tier two, where AI starts to get a little risky.
Today, we are finishing tier two, going through tier three, and closing with the three-question framework you can use every single time going forward.
But first, I want to add something that came up after part one dropped, and I think this is important to share. As I was reflecting on it, I realized there was a use case I did not cover that I think a lot of you will use.
I was thinking back to athletes working with us and when they come in, what are the things they share that they are struggling with? I think it is one of the most valuable things AI can actually do in this process, but it is also one of the trickiest things to get right.
This is where the audit use case comes in.
Something I have been thinking about since part one is using AI as an audit tool. I think it applies to all three audiences: the ACL athlete, the parent or support person, and the clinician or coach.
Here is what I mean.
You are going through rehab. Maybe things feel off. Maybe your gut is telling you something is not right, but you cannot articulate it. Maybe your surgeon said something that did not fully land, like you are three months out and they said to start running, or you are six months out and they cleared you for sport. Maybe the PT gave you guidance, you are just not sure about.
It could be those same things: clearing you to run at three months without testing, or saying knee extensions are very bad while you are seeing all this information about knee extensions being amazing. Side note: they are very useful in rehab, so whoever says otherwise is not really up to date.
You are not trying to replace their judgment in these situations. You are trying to understand it better and figure out if the questions in your head are worth asking.
It is a way to use AI as a soundboard, not to make the decision itself. That is important here. It is to help you audit the process against what the research and science actually say.
I even think about this podcast. Yes, there are a lot of things I share my opinion about. I share what this landscape of ACL rehab looks like, what return to performance and sport looks like, what re-injury rates look like, and the nuances in this process.
But I try my best to pull not just from anecdote. Yes, I share stories from you guys, but if you look at our foundation, it is rooted in research. It is rooted in science. It is rooted in data.
That is where we need to operate from.
It is not just, “I had a hamstring graft, therefore hamstring grafts are better.” Or, “I did this one type of therapy, and that got me back.” Or, “I did this one exercise, and it was the magic bullet.”
It is not about methods first. It is principles first.
We distill things down to the core basics. Physics is physics. We are not changing physics. We use principles to guide exercise selection, constraints, and the way we navigate this process.
We use different models. None of them are perfect, but they help guide foundational principles for understanding how to rebuild you from the ground up and get you back to being an athlete and being active.
That is what matters most.
Having a soundboard is helpful, and AI can help process through that instead of making the ultimate decisions.
Here are some examples.
You ask AI: What does the research say about criteria for starting running after ACL reconstruction?
Not what your PT says. Not what your surgeon says. What does the research say? What is the aggregate of studies that have looked at controls, experimental groups, biomechanics, psychology, lifestyle factors, strength factors, anatomy, and all the variables that contribute to successful or unsuccessful outcomes?
Then you compare that to what you are actually being told.
This is one of the biggest dilemmas ACLers face. I hear so many people say, “My PT says this,” or “My surgeon says this,” and they are not sure what to believe. It is okay to be unsure.
The key is becoming equipped with more education and understanding so you can go with what they say, push back on what they say, or ask better questions.
That is what this podcast exists for. It is not to replace anything. It is to help inform you in making better decisions.
ACL injuries are at an all-time high, and return to sport and performance is at an all-time low. It keeps getting worse in both directions.
Even with all the research and information on this injury, we are still not getting there yet.
The thing I keep thinking about is that the information already exists.
There is not going to be some magical study showing a secret exercise that changes everything. We know that good rehab, strength and conditioning, getting your quad stronger, and managing the key variables are what matter.
We know the basics work.
Testing exists. If we did more evidence-based testing, we would lower re-injury rates and get more people back to the things they love.
A lot of this comes down to education from surgeons and physical therapists, the process and frameworks they use, and timeline-based protocols that are not helping ACLers.
A lot also comes down to insurance and healthcare limitations. The model itself often works against ACLers because it is not designed for this rehab. That is why rooting things in research and science matters.
Another example:
You ask, what does proper quad strength testing look like after ACL reconstruction, and what are accepted benchmarks before return to sport?
If you ask your PT or surgeon this early on, that is part of vetting the process.
There are a ton of PTs and surgeons who will say, “We’ll check that when we get there.” They avoid the question because they cannot answer it or they do not do it.
Point blank: if your surgeon cannot tell you what they are looking for, and especially if your physical therapist cannot explain strength testing clearly, that is a problem.
They should be able to say:
This is what strength testing looks like.
We are using an isokinetic machine.
We are using isometric testing.
We are using a 3-to-5 rep max method.
These are the benchmarks.
This is when we test.
This is how often we re-test.
These things need to be explicitly stated.
Every one of our athletes knows these KPIs and benchmarks. We revisit them constantly because it lets athletes know where they are and what they are working toward.
You can ask AI what proper strength testing looks like after ACL reconstruction. Then ask yourself whether anyone has actually tested you that way.
Do not be surprised if they have not.
Honestly, I would estimate 80 to 90 percent of physical therapists and surgeons are not properly testing ACLers. That means maybe 10 to 20 percent are, and that is generous.
That is the reality.
Another example:
Ask what the evidence says for return to sport after ACL reconstruction.
You may get a list like strength symmetry, hop testing, vertical hop testing, psychological readiness, graded sport-specific exposure, power performance, range of motion, knee symptom profile, and more.
Then ask yourself: Have any of those boxes actually been checked in my care?
That is a legitimate use of AI.
Use research and science as your foundation, not opinion or anecdote, and hold your care to a proper standard. Because right now, the bar is on the floor.
We need to pick it up. We need to raise the standard. You deserve better.
Now here is the critical thing.
AI is a “yes man.” That is not an exaggeration. It often gives people what they want to hear. It shapes responses based on how you ask the question. “Garbage in, garbage out.”
If you ask in a leading way, it can validate the answer you already wanted. If you ask neutrally, you may get something more honest. Sometimes it helps to tell it to play devil’s advocate or be contrarian so you can think from both sides. If you are going to use AI to audit your care, your prompt needs to be as neutral as possible.
Do not ask: My PT has not tested my quad strength yet. Is that bad?
That is a leading question.
Instead, ask: What does the research say about when and how quad strength should be assessed during ACL rehabilitation?
Now you are asking for science, not validation. That is where AI has the most to offer.
The moment you ask it to evaluate your specific PT, your specific surgeon, or your exact case, you are entering territory where it lacks enough context to be reliable.
Use it to get educated. Use it to build better questions. Use it to walk into your next appointment knowing what the evidence says so you can have a real conversation with the human responsible for your care.
Done right, the audit is powerful.
Done wrong, it is just another way to find the answer you already wanted.
Now, picking up where part one left off. We talked about the confidence gap and confirmation bias—the athlete who asks AI the same question in different ways until they get the answer they want.
I want to add one more pattern before we move into tier three.
The athlete who builds a program from AI and then uses AI to troubleshoot it.
No human in the loop. No one watching them move. No one tracking their quad. No one is catching compensations. No one noticing swelling patterns.
AI does not accumulate knowledge about you between sessions the way a real clinician or coach does. It resets. It misses nuance. It rarely says, “I do not have enough information.”
That is where things go wrong quietly over time.
Do not be that person.
Tier three asks this question:
Does this require someone who can actually assess me, watch me move, and take responsibility for the outcome?
If yes, AI is not your answer.
Building your ACL rehab program is one example.
AI can create something that looks like a program: phases, exercises, sets, reps, progressions.
That is the problem.
It looks legit.
But it does not truly know your graft type, tissue healing timelines, quad deficit, swelling patterns, movement quality, concomitant injuries, sport demands, or psychological readiness.
It gives you the average of generic ACL content on the internet—which is often protocol-driven and timeline-based.
That already fails athletes.
If you are using AI to build a program because you had no other options, I get it.
Maybe insurance ran out. Maybe care was poor. Maybe resources are limited.
That is real.
But you deserve more than a generic template from a language model.
There are in-person options. There are remote coaching options. We do that. Others do that too.
This exact gap is why we built our services—to help ACLers who would otherwise fall through the cracks.
Another example: the return-to-sport decision.
Return to sport is criteria-based.
We look at quad strength symmetry, hop performance, deceleration mechanics, movement quality under load, power output, jumping, cutting, agility, psychological readiness, and graded sport exposure.
AI can tell you the list.
It cannot tell you whether you meet it.
Those are very different things.
If an athlete asks AI whether they are ready and gets a permissive answer, they were not cleared.
They had a conversation with a chatbot.
ACL re-injury rates for athletes returning to pivoting sports are often around 20 to 30 percent.
That number does not come down by skipping testing.
It comes down by doing it.
The ongoing coaching relationship matters too.
A real human accumulates knowledge about you over time.
They know your history, tendencies, patterns, and changes week to week.
AI cannot replicate what a good PT or coach does over months.
These are not comparable things. And this goes beyond exercises.
People are dealing with family loss, school stress, breakups, fear, parenthood, transportation issues, identity disruption, and life stress while also trying to rehab an ACL.
That context matters.
The mom who cannot carry her six-month-old because she is non-weight-bearing.
The athlete who cannot drive because the surgical leg is the driving leg.
The person spiraling mentally because their knee felt weird for three days.
We work with the whole person, not just the knee.
That matters.
Psychological support is another huge gap. AI can explain kinesiophobia, the fear of movement. It can normalize fear of re-injury.
But can it work through grief, identity loss, or fear that lives in your body when you try to push again?
No. That requires a real human with real expertise. Sports psychology is one of the most underutilized parts of this process.
Now let me share how we use AI at the ACL Athlete. We are not anti-AI. We are tech-forward. We use technology constantly. Used correctly, AI can help us be better for our athletes. But it does not replace what we do.
We use it for research deep dives, reviewing evidence faster, analyzing testing data, building dashboards, translation of reports from other languages, parsing complex imaging notes, and summarizing coaching sessions into clean takeaways and action items.
It can also help us find providers in cities where we have no network. But here is what it does not do: It does not build programming. It does not make return-to-sport decisions. It does not replace the coaching relationship. That relationship is the core.
Could we automate more? Sure.
Would it save time? Sure.
Would it remove what matters most? Absolutely.
We care deeply about human-to-human connection.
Knowing the athlete. Knowing their kids. Knowing their sport. Knowing their dog. Knowing their stress.
That week-over-week understanding is irreplaceable.
AI should improve the surrounding work so we have more capacity to do the important work better. That is the goal.
People ask if I am scared AI will replace what we do. No concerns whatsoever. Not because of ego.
Because I know how much nuance, detail, judgment, empathy, and responsibility this process requires.
I have lived it personally. I have had two ACL injuries and many others. I have had thousands of conversations with ACLers. This work is deeper than people realize.
What I do think AI will replace is generic rehab. More people will be pushed into apps and automated home systems. We are already seeing it. But ACL rehab requires nuance. It needs to be dialed in.
Now to close this out, here are three questions to use every time.
One: Am I using this to get smarter and more prepared, or am I using it to make a clinical decision? Getting smarter is what AI is for. If you are asking it to decide something requiring clinical assessment, you crossed the line.
Two: If AI gets this wrong, what is the actual consequence Wrong about a term? Your PT can clarify it. Wrong about being ready for sport? You re-tear at nine months. The consequence tells you how much human expertise needs to be involved.
Three: Does this require someone who can actually assess me, watch me move, and take responsibility for the outcome? If yes, that is a human job. AI can support it. AI should not do it.
If you are an athlete right now, you deserve someone who knows you, tests you, watches you move, and builds your program around your actual data and your individual needs.
AI can help you find that person and show up prepared to work with them. It cannot be that person.
If you are a parent or support person, use AI to get educated and ask better questions. If something feels wrong, seek another qualified human opinion.
If you are a clinician or coach, AI is leverage. The filter is you. If you remove yourself from the filter, you are asking it to do your job.
AI is a powerful tool, y’all. And it is a powerful crutch. I’m going to say that again because it is good. AI is a powerful tool, and it is a powerful crutch. The difference is whether you are using it to show up better for the humans in your corner, or trying to use it to replace them.
If you want an actual human in your corner from surgery through return to sport and beyond, you can reach out to us.
We are going to talk more openly about working with people because I realize many of you do not know that we offer remote coaching and work with ACLers all over the world.
If you need help, please find us at theaclathlete.com.
You can also look in the show notes for one-on-one remote coaching.
We work with many sports, ages, and levels.
A lot of people are skeptical of the remote process, which is why we offer a free consultation so we can decide together if it is a good fit.
If it is not, I promise you will still leave with the next steps and clarity.
Whether it is with us or elsewhere, I want you to have more power in your story and more clarity in your next move.
What I want from you guys is to share this series with someone who needs it—another athlete, a parent, or a clinician.
This conversation matters so much, especially for the future ahead.
See you in the next episode. This is your host, Ravi Patel, signing off.
Subscribe and leave The ACL Podcast a review – this helps us spread the word and continue to reach more ACLers, healthcare professionals, and more. The goal is to redefine ACL rehab and elevate the standard of care.
Resources:
- Check out our free ebooks on our Resources page
- Sign up for The ACL Athlete – VALUE Newsletter – an exclusive newsletter packed with value – ACL advice, go-to exercises, ACL research reviews, athlete wins, frameworks we use, mindset coaching, blog articles, podcast episodes, and pre-launch access to some exciting projects we have lined up
- 1-on-1 Remote ACL Coaching – Objective testing. An individualized game plan. Endless support and guidance. From anywhere in the world.
- More podcasts? Check out our archives
Connect:
- Have questions or a podcast idea? Send us a message
