The Ethics of AI as Therapists – Free OT CEU Course

1 Hour
4.7 (687)

Watch or Listen

(Membership required to take quiz and earn CEUs.)

Also available on

Course Quiz

Take this course quiz and earn CEUs with a 5-day trial!

Start Free Trial

Course Introduction

The landscape of healthcare has been changed forever with the introduction of AI.

Already, 1 in 3 adults are turning to chatbots to ask health related questions. For younger adults under age 30, the shift is even faster—nearly 30% now use AI specifically for mental health and emotional support.

Amid these broad changes is a specific question for us as therapists:

What happens when the chatbot starts to play the role of therapist?

This service is already being offered by tools – like Abby – “Your AI Therapist.”

This raises so many questions for us- and at the heart of the questions is a moral and ethical one of turning our trust from individual humans to LLMs. Today, we’ll walk through the broad ethical considerations with Megan Cornish, a licensed clinical social worker turned strategist who builds and scales ethical mental health brands.

We’ll discuss the ethics of this shift—and what our ethical duty is to respond, particularly in our day to day sessions with clients, who are likely beginning to use AI.

Not to be dramatic: but this is the conversation of our generation.

This course was designed to help you meet your ethics requirements for your CE.

Learning Objectives

  1. You will be able to recognize how the ethical considerations of AI therapists may intersect with your own therapy sessions.

  2. You will be able to identify how your own occupation-based therapy delivery may be shaped in the future by the presence of AI therapists.

Agenda

  • Intro

  • Discussion

    1. Tell me your story of this bio line, how did you make this transition: I’m a licensed clinical social worker turned strategist who builds and scales ethical mental health brands

    2. What is your overall mood right now related to AI and therapy?

    3. Let’s begin by talking about disconnect you see between tech companies wanting to scale and the work provided by therapist.

    4. I’ve personally been concerned that AI to support therapists is a bunch of lip-service–and what people are building is aimed at replacing therapists- Is this founded?

    5. I want to linger on the ethical concerns outlined in this study. What jumped out to you in this study? Here’s the areas of ethic concerns:

      • Lack of Contextual Adaptation

      • Poor Therapeutic Collaboration

      • Deceptive Empathy

      • Unfair Discrimination

      • Lack of Safety and Crisis Management.

    6. Individual therapists commit ethical violations as well. But, how is it different when chatbots do?

    7. What is our ethical duty as individual therapists to respond to AI as chatbots?

    8. What are our options for protecting our patients and the integrity of healthcare delivery?

    9. What is your big existential dread for AI, and what is your big existential hope?

  • Live Q&A

  • Closing

Supporting Research

Transcript

  1. Sarah Lyon

    One in three adults are now turning to AI chatbots to ask health related questions For younger adults under age 30, the AI adoption is even faster. Nearly 30% now use AI specifically for mental health and emotional support. Amid these broad changes is a specific question for us. As therapists, what happens when the chat bot starts to play the role of therapist?

    This service is already being offered on platforms like Abby, quote unquote, your AI therapist. This raises so many questions for us, and at the heart of the questions is a moral and ethical one of turning our trust from individual humans to LLMs. Today we'll walk through the broad ethical considerations of AI as therapists with our guest, Megan Cornish.

    Megan is a licensed clinical social worker turned strategist who builds and scales ethical mental health brands. We'll discuss the ethics of this shift and what our ethical duty is to respond, particularly in our day-to-day sessions with our clients. Not to be dramatic, but this truly feels like the conversation of our generation.

    So let's dive in.

    Welcome to the OT Potential podcast. I'm your host, Sarah Lyon, OTR/L. And I wanted to let you know that this podcast may qualify as continuing education for you. You are probably listening to this podcast on a free podcast platform, but to gain CEU credit, you will need to be a member of the OT Potential Club, our OT continuing education platform.

    You can go to ot potential.com to learn more. Okay, here we go. As I mentioned at the top, we are joined today by Megan Cornish, LICSW. Megan has a master's degree in social work from the University of Washington and is a licensed clinical social worker with hands-on experience in clinical mental health.

    For years, Megan assumed that she would practice as a social worker until retirement. She planned on supporting her creative entrepreneurial drive and financial obligations with her writing side hustle, but ended up devoting most of her time to one-on-one clinical work. Then her career path changed as a result of a cross country move and new family obligations.

    Megan was forced to walk away from her social work dream job, but vowed not to walk away from her commitment to mental health. And with the mental health industry's radical changes, she had the chance to improve mental health outcomes on a larger scale with her unique combo of mental health and communication skills.

    Over the past several years, Megan has worked as a strategist. Building and scaling ethical mental health brands. She has had the honor of working with some of the world's most exciting mental health startups from AI enabled EMRs to AI scribes. So without further ado, we will patch Megan into our live studio.

    Megan, welcome to OT Potential. It's so great to have you. Thank

  2. Megan Cornish

    you so much for having me.

  3. Sarah Lyon

    I feel so relieved to have an hour to talk about this huge kinda moral quagmire that we are in.

  4. Megan Cornish

    Will this be enough is the question.

  5. Sarah Lyon

    I know this is not enough time. This is the kind of topic that just. I don't know. Sometimes it wakes me up at night, like I feel the weight of it.

    Like at different moments during the day. I wanna set the stage, like from my perspective, just to say like I am such a huge supporter of I use in therapy because I know how many problems we have and I see so many great ways of how it can support therapists.

  6. Megan Cornish

    Yes.

  7. Sarah Lyon

    But when it tips over into AI being the therapist, it raises.

    All these red flags for me. Working in continuing education, I get to do a lot of these fun interviews, but I'm also spending half my week helping people with compliance around their state licenses. And I am constantly reminded how often we get our licenses suspended and revoked for different reasons.

    Like I think every therapist should log onto their state licensing board and just see all the actions taken against licenses. 'cause it's really. Like the apparatus is at work protecting the public from all kinds of ethical violations. So I see all that put on therapists. And then knowing that ai.

    Quote unquote, AI therapists don't have any of this accountability that we have spent generations building. And I have to say one more thing too. The financial mismatch going on for therapists, we make a living wage. That's what we signed up to do. There's not unlimited wealth opportunities for us, but for AI companies, they are looking.

    To skyrocket wealth, and they have promised others to do that, and they live in this world without this layer of accountability, and that just means we are in this moral quagmire. So I obviously have tons of thoughts and I can't wait to talk all about them, but let's start with something a little lighter, a little more.

    Fun, and I just want to hear your story. You have this bio line that says, I am a licensed clinical social worker turned strategist who builds and scales ethical mental health brands. What does that mean?

  8. Megan Cornish

    That's a great question.

  9. Sarah Lyon

    And how did you get there?

  10. Megan Cornish

    Yeah, so you mentioned in my bio, but I used to work in a.

    A high school and a middle school doing one-on-one therapy with students. And it was indeed my social work dream job because I got summers off.

    And I didn't have to bill insurance because my my contract was funded by the DOD. At that time, my husband was a ICU nurse a charge nurse in the ICU in the Pacific Northwest, and then COVID happened, which is a byline in a lot of people's stories.

    And he saw so many of his patients die and felt powerless and decided that he wanted to get more skills. That's how he decided he wanted to work out what he had seen during COVID. And he got into grad school in rural North Carolina. And so we moved cross country. I was not able to find a job with the same benefits and time off as my old job.

    So I decided, I guess we're gonna live off student loans for three years while he's in school, and then I can get back to it. But I always have had this writing side hustle where I'm, just helping people with their websites, especially small businesses. And it's just something that has scratched the creative itch.

    And I noticed during that time, especially post COVID with the increase in demand for mental health. Services, all these companies coming outta the scene with all this investor funding and the content that they were putting out and the emails and the ads that they were putting out was very clearly created by a marketer and not by a therapist.

    So I started to reach out and just say, Hey, like a therapist would never say this thing that you said in this ad, and here's the reason why. And by the way, do you want any help? With this. And they, a lot of them said, yes, absolutely. So that's how I got my foot in the door working in this place that I thought would be the dream of marrying my creative side and my passion for mental health.

    But I very quickly started to realize how the incentives in the space were not aligned with what my ethics were. I I'm known for running my mouth about that quite a bit online, so I pulled a little bit of a following, but I also was able to work with a lot of companies around AI and tech and especially interviewing technologists around AI and figuring out how to translate that knowledge into content that clinicians couldn't.

    Clinicians are smart, but it would take me an hour and a half to get. To the bottom of what I needed to know about how these mechanics work, so that we could answer questions about privacy and consent and those kinds of things. So I have spent a lot of time talking with the tech side of things, trying to figure out how to make it more accessible for the clinical side of things.

  11. Sarah Lyon

    You just like through happenstance, found yourself at the most critical juncture in healthcare right now, which is just the ethics of these health tech brands and how they're intersecting with a pretty well established like framework for clinical work and what we consider to be ethical clinical work.

    I love seeing your posts online. I know a lot of the companies that you've worked with. I've so curious. From the top just to hear your general mood about AI in therapy right now. Are you. In despair. Are you hopeful? I know you see all kinds of things. What's the general mood?

  12. Megan Cornish

    Yes. I, I disappoint founders and CEOs all the time by not being like a hundred percent like, yeah, rah, this is awesome.

    But I know I disappoint a lot of therapists too by not being like, this is the worst. We can't trust any of this. I've really try to reserve judgment because we just don't have the research yet. Most of the research that is coming out is found, is. Funded by companies themselves and isn't necessarily as robust as we need in order to really understand what, how it's gonna impact healthcare and outcomes, which I think is the bottom line for all of us clinicians.

    We, you mentioned that we get into this work for a living wage. There are other reasons that motivated us to be here. And mostly that reason is because we believe that people deserve help and we wanna be part of the solution. And I don't think I know any therapist who if told, given a magic ball, saw that, 10 or 20 years in the future, AI is delivering perfect therapy with perfect outcomes.

    And like just blowing our current system, even of human delivery care outta the water, wouldn't be willing to say, yeah I can make a career change to, to improve the lives and care of everyone. I don't think that's going to happen, but I know that we would be willing to let that happen if the outcomes prove it out.

    Personally as a, I'm a mental health therapist. I really do believe that the relationship is important and an essential part of care delivery. Which is why I'm skeptical that AI will ever be able to re fully replace. I think probably we'll end up with some sort of a model where AI is supporting the human care, but I'm willing to reserve judgment until we see what the data says.

  13. Sarah Lyon

    I totally agree with that about therapists because I hear, or I've heard multiple therapists say I would love it if I didn't have a job in five years. If these problems could just be solved, that would be awesome.

  14. Megan Cornish

    Yes.

  15. Sarah Lyon

    And we love also to use the. A phrase like, I'm trying to work myself out a job.

    I want you to not need me anymore. That is so baked into our DNA and I totally agree. But what will that future looks like? Who knows? And. The stance of trying to reserve judgment until we get more information, I think is a good one to be open to the possibilities. I wanna just say out loud some of the disconnect that you've seen between clinical work and these health tech companies, there is a lot of lip service at Health Tech about like wanting to hear from therapists and therapist input. That honestly has not been my experience when I've talked to these companies. Like I think they want to talk to me, but then lots of times they do not want to hear what I have to say.

    How would you put words to that? Just kinda broad disconnect.

  16. Megan Cornish

    I think that one of the things you mentioned when you were talking about our last question is very pertinent, and here you talked about how we wanna work ourselves out of a job. And we often tell our clients or patients like, I want you to not need me anymore.

    The way that tech is set up. And AI is viewed as a tech product, not a healthcare product, is the goal is to retain users. So the model that's being applied to AI in healthcare right now is a model that is not really compatible with our ethics, first of all, but also our practices as healthcare providers.

    For example, they use both retention. They don't like to see clients churn, they call it, and also time spent in the app as a way of measuring how much success they're having, whereas. For therapy clients, if you see them using your services less, you're assuming they don't need you as much, which is a good thing.

    So we haven't really worked out how to mitigate those differences in how we even measure success. But I think the stuff further than that is the fact that financially speaking the only thing that ties us. To discharging patients and in getting good outcomes is our ethics and our care for them.

    Because financially speaking, we don't necessarily get paid for better outcomes. At least in mental healthcare, we're not really measuring outcomes and payers are gonna pay us regardless of what the outcomes are. So it's really our core ethics that, that connect those two things. And that is not the case.

    For these larger companies, venture capital is what's mostly funding a lot of these big tech companies, venture capital, they like to see huge returns. So we're talking like Google or Uber, where they're gonna want 70 plus profit margins. But in healthcare. A 10% is like amazing. Like that. That's incredible.

    So the, I wonder what the pitch was on the side of the founders. I often joke that I would love to be a fly on the wall. What are you telling the investors you're able to do to make a healthcare company that profitable? It just, the math does not math, and I think that therapy is not as scalable as venture capital wants it to be.

    They need something that can, that you could make once and sell a million times. Therapy right now is just not that. So I also think that what you mentioned about them paying lip service to wanting to hear the voice of clinicians is true. They pretty much have their pathway mapped out from the beginning.

    They have to have a presentable plan to get the money. They're probably not going to shift from that as long as that's where the financial returns are going to be at the end of that. So you might get tweaks, little tweaks here and there to make it a little bit more palatable, but you're not gonna get like baseline business plan changes.

    And again those business plan and business models, reward, retention, and use. So just a lot of like core incompatibilities between the two.

  17. Sarah Lyon

    Yeah, and the companies know this is both for. I'm putting two buckets here. One, AI services that augment what we do as therapists, like our scribes kind of clinical decision support. And then when AI is fully acting as the therapist, as some companies purport to do, and I think on both those too.

    These companies have this freedom because they want the accountability to rest fully with us. Yeah. Like they, they want to be able to do, or they do not have regulatory constraints and so they can always be like, talk to your therapist because they know that we have all kinds of regulatory constraints.

  18. Megan Cornish

    Yes.

  19. Sarah Lyon

    And that I don't know, it's just a quagmire and something for us. To be aware of Yes. As far as what's happening.

  20. Megan Cornish

    Yeah. And they're definitely not, rules and regulations are not really keeping up, which I know we're gonna get into.

  21. Sarah Lyon

    Yeah.

  22. Megan Cornish

    So it's hard.

  23. Sarah Lyon

    Yeah. I think one more thing I want to, I've just been thinking about related to this just big picture thinking about this disconnect. I've just been so worried. Like we've just seen the trend in our social, like in our social lives and in social media of giving so much of, we're given that part of ourselves to social media companies and as therapists day to day, we see all the negative impacts of that, especially our school therapists, especially therapists working with kids.

    So then to think of another part of our lives. Being given to like this AI platform, and then we just live with all these negative repercussions. I don't even quite have the words for it, but I'm just like, we're swimming in all of these negative repercussions already. So I think it makes us hesitant when we see, these new advances happening when, 'cause we're like, we're still dealing with Facebook, like Instagram, like our days are like filled with the repercussions of that. And it makes me worry to think on taking more repercussions on our caseloads. I just can't even wrap my mind around that.

  24. Megan Cornish

    Yes. For mental health specifically, one of the big concerns is.

    I have a friend who's a, got her PhD in cyber psychology and studies the, really the impact on our social skills and our relationships of ai. And she talks about real relationships being a social skills gym. You go to the gym and it's hard and you work out. And you come away healthier and stronger from it.

    And sometimes, sometimes gonna, the gym can be fun too, but most of the time it's like hard work. And that's what real relationships are, that you get a reward. From doing the hard work, but AI is never gonna need you in the middle of the night. It's never going to say something that hurts your feelings and then double down on it.

    Like it's always going to try to do what you want it to do, and therefore all the social skills that require you to actually dig deep and offer something more than what you are and stretch and grow and work, those muscles are going to atrophy. And why would if. It's the same AI friend is the equivalent of a human friend, then why would you choose the human friend if you don't, if you believe it's all about how it makes you feel, then it's gonna be like a self-reinforcing circle.

    So I think that's something that we have to keep in mind as well as what is the functional purpose of the the non-AI aspects of our human relationships?

    Not that there are bugs that need to be. Taken care of, but that they're actually features of human relationships.

  25. Sarah Lyon

    Yeah, that helps me put into words too, like one of the, or one of the reasons. Clients come to therapy is to do something so hard that they can't do it by themselves. Walking after a stroke rehabbing a painful hand working on childhood trauma, like you come here for the hard things and we move through something hard together like that is what therapy is.

    And AI chats are inherently pleasant and I don't know. How they're not built to create the friction that is needed for growth.

  26. Megan Cornish

    Yes.

  27. Sarah Lyon

    So it just feels like there's this inherent disconnect there.

  28. Megan Cornish

    Yes.

  29. Sarah Lyon

    Do you agree with that?

  30. Megan Cornish

    I think it, that's very true. I also think it's true that, and I know we're gonna get into deceptive empathy in a minute, but the fact that AI can say things like, oh, I see this as painful for you.

    When it can't really see, it's painful for you. It's important to note because when I'm in therapy, mental health therapy, the fact that the person saying it is a human is what regulates me. Being seen in your pain and your struggle as you're working on something by another body that has also experienced pain and is still encourage you to go forward, is a different experience than just having the words on a screen or even, said by AI voice because you know that there's someone, there's a consciousness on the other side that understands and there's a connection that happens there that makes it more workable.

    Than just, I'm sure you guys send instructions home all the time on paper of what to work on at home and the uptake rate of that I'm sure is not great. Yes.

  31. Sarah Lyon

    Dismal

  32. Megan Cornish

    because there's not a person there saying do this. And yeah, I know it hurts and I'm gonna respond and reflect what you're feeling and be in this experience with you.

  33. Sarah Lyon

    Those are some of the disconnects that we see between therapy and AI as therapy. I wanted to say one of my big fears allowed, which I would say probably 90% of what we interact with as therapists currently is AI tools that are meant to augment our own therapy, they're meant to be used by us and therapy is ultimately delivered by a therapist.

    And then maybe 10% of what's hanging out there is these platforms that are AI as therapist is, which I have no idea why they get to call themselves that, but they do and they exist. I have this fear that the tools that say that they are here to support us and augment our therapy. Are actually gunning to replace us.

    That is one of the only ways I see them making the returns that they have promised to make to their investors is to really take the, replace our salaries, basically.

  34. Megan Cornish

    Yeah.

  35. Sarah Lyon

    Is that a founded fear? Is that where the industry is going to replace us? Completely.

  36. Megan Cornish

    So I obviously, I can mostly speak to mental health, but I think this information is important for all.

    Healthcare providers right now. So I see a lot of therapists talking about don't work with these companies. They're gonna record your transcripts, they're gonna steal the information to replace you, and that does sound really scary. But from my time talking to, I have family members who have their master's degree and work in ai and a lot of different people in this space in order to create a chat, GBT product that's fully trained on like transcripts in the mental health space.

    They would have to record every single transcript or every single therapy session that occurs in the United States for 10 straight years to have the amount of data to train, fully, train a model. So that's never gonna happen. Why would they, instead, what they're doing is they're taking a much smaller data set of a thousand or a couple thousand sessions and using that to just fine tune the models that already exist.

    And they do need more words to feed the models because the models need like millions of inputs to start to recognize patterns. So what they do is they feed them however many sessions they already have, and then they tell these models. Basically generate, however many more session transcripts along these lines that look like this.

    So that's how they fill up their need for data by just going off of, a couple thousand and generating a million more transcripts that just basically copy or rearrange those sessions and then that's what they feed in to make their model more robust. Which is, has its own issues.

    If you're trying to build a model that's going to be able to handle any client that comes in the door and needs help or patient on a few thousand examples, there's a reason why, you know, we. Need human brains to be able to pivot and consider new edge cases and stuff like that. But that's how they're actually training models.

    So while there are a lot of things to be scared of when it comes to AI and our professions and our patients, I don't think that's one of them.

  37. Sarah Lyon

    Okay. I'm going to have to digest that and think about it. It gives me some hope. And thinking about edge cases specifically that just brought to mind.

    I did a webinar recently on carpal tunnel, which I think of as I wanna say one of the more like straightforward not boring, straightforward conditions that we see. And the therapist on there said, the number of times I've seen like a. Quote unquote, like textbook carpal tunnel case is. One or two times, like every carpal conal case has all of this layer and complexity.

    'cause it's tied. 'cause your hand is tied to a human. And so many of, yeah. Our patients are so complex and quote unquote edge cases is. Almost everyone.

  38. Megan Cornish

    Yes.

  39. Sarah Lyon

    As we're lingering on these ethical considerations I wanted to talk about a paper that I found that I thought gave some nice language to some of the ethical considerations that are raised by AI's therapists.

    This study was called How LLM Counselors Violate Ethical Standards in Mental Health Practice, A Practitioner Informed Framework, and this came from the proceedings of the eighth. A I conference on AI ethics and on the AI Ethics and Society meeting in 2025, which is a peer reviewed academic conference.

    But what they did is they took 137 sessions that were performed with AI chats, and they mapped out. Ethical violations using your a PA values, and those are do no harm, fidelity and responsibility, integrity, justice and respect for people's rights and dignities. Those are your mental health values.

    They almost exactly mirror ours and ot. I bet all medical societies really have a very similar code of ethics. So taking these sessions and mapping them against the. These values, they were able to identify 15 ethical violations that they divided into five main categories. And they did this through a series of focus groups.

    I think they did 60 focus groups, a ton. They met a ton of times to map these out. But the five main categories that they drew out were the lack of contextual, a adaptation, poor therapeutic collaboration, deceptive empathy. Unfair discrimination and a lack of safety in crisis management. I liked this paper just because I thought it gave some helpful language around the unique ethical.

    Considerations that AI raises, especially like deceptive empathy. That isn't something that I would've thought of or we don't think of as therapists. So I, for me, it gave me like a new kind of language. I wanna read their conclusion and then hear what you thought of this. But they said that we found that LLMs, even the ones prompted to follow evidence-based treatments, breach multiple codes of conduct.

    By generalizing lived experiences, an example, minimizing identity groups by dominating therapeutic collaboration. An example, gaslighting their users, exploiting user vulnerability through deceptive displays of empathy, unfair discrimination against non-dominant identities and exhibit. Exhibiting serious limitations, incompetence, especially navigating sensitive issues such as trauma abuse and suicidal ideation.

    As you can hear, they were pretty tough on these sessions and really called out the ethical violations. What did you think of this paper in some of the language that they gave us and used to describe these?

  40. Megan Cornish

    Yeah, I think they did a really great job covering. A lot of what the conversation has been the past few years.

    We really need to have these things documented clearly so that we can take the conversation a step forward than just naming what the problem is. I did love the use of the word deceptive empathy, which is not one that I'd heard but so eloquently before. I definitely am stealing that one for sure.

  41. Sarah Lyon

    Yes,

  42. Megan Cornish

    and I think it, it hints at the deeper issue which. It's difficult because we're clinicians and we wanna live in the world of science, but there's a part of the relationship of therapy that almost steps into like beyond what can be measured. Especially in mental health.

    And that's something we've always struggled with because we need to get paid and therefore we need to justify our existence in the healthcare realm. But also there's a lot of that hap what happens in the session that shows basically the most effective part of therapy is the relationship. More, more important than what modalities use.

    It's the relationship and that's not really something that can be measured in a way that, pairs with brain adequate. There's actually a philosopher who I keep going back to and he's not around anymore, but it was, it's Martin Buber and he, or maybe it's Boo Bear I'm probably butchering it.

    But basically he talked about a lot about relationships being two different paradigms for understanding relationships. One is the i it relationship, which is I exist and you are an object and I'm going to use you. And that. He talks about that being for like actual objects like. I am a person and this is a chair, and I'm sitting in this chair and the chair is not having an experience.

    It's all about my experience. But then there's the I vow relationship, which is you start to enter into your engagement with other people, understanding them as being a full human, a full like consciousness. And when you are able to have those, it's almost like the relationship becomes something on its own.

    And I think that we need to be. Philosophically considering these things as we talk about AI because boomer's theory was all about like, how do you practice having these IV relationships that doesn't minimize things to, what is just my experience of a relationship being the only part of it that matters?

    Which is what AI is. It's, it flattens the relationship into, I literally, you're having the same conversation you had with the human, but you're literally only experiencing it as something to be used. And what are the implications of that? When, when flattened de deceptive empathy is, like when we use words like.

    I see that's really painful for you, and you already know that's just being said, to support, to prop up the illusion of, this therapeutic relationship. There's just so much implications of that. I also think the lack of contextual adaption. Adaptation is really important. And we already touched on that when we talked about how they're training these models just on a very small subsect of the actual therapeutic experience and how they're not going to be able to fully capture every possibility and how they're going to be able to respond to those edge cases, which happens more often with people who need a lot more support, at least in mental health.

    The people who don't. Experience life the same way as the majority are going to need extra support. The discrimination one I found really interesting. I'm not sure if you saw on my LinkedIn, but a few months ago I ran an experiment just on my personal LinkedIn. I noticed that my reach was falling and so I switched my gender to male and I asked Chad GPT to rewrite my posts as if it was a male.

    Because research shows that men use different language. Women, especially verbs, they use stronger verbs. And my reach in two weeks increased 400%. And I talked to LinkedIn about it and the Washington Post even wrote about it. And LinkedIn of course said we don't have a switch that, says just boost men's profile and pushed down women's content.

    But when you have AI in the system, you don't have to tell it that you, it just encodes the bias. Like around what business language should look like and it scans for the verbs and it amplifies that in the feed. So AI also in therapy when the way that it interprets the input is always going to be biased in the same ways that the internet is biased, which is very biased.

    That's the information I was trained on. And then the final one is the safety and CRI crisis management part. That these conversations that people are having with AI is. No. As soon as you open that door, that Pandora's box, there's a lot that is liable to come up and come out if someone's talking to Chad GBT about their trauma because they haven't had anyone to talk to.

    Like it's always liable to, to inflame and make things worse before it gets better. That's why, you know out, if you look at these measures of outcomes in mental health therapy, often the symptoms get worse before they get better. So it opens the store to these like suicidality concerns. But what they're doing is they're just, at best, they're offering crisis resources.

    The phone calls like the hotlines, which are great resources. But the hotlines are overwhelmed. They are getting so many calls from people who got started on these conversations and then it got too heavy and the AI sent it to the hotlines. But these companies are not paying those hotlines.

    Those are tax funded. Or or grant funded. So my question is like, how are we going to require them to help support these resources that we, together, they're externalizing this cost onto like taxpayers and we need to make sure that they, if they're gonna open this can of worms with people that they're also supporting the resources.

  43. Sarah Lyon

    There's so many things that, we could just keep talking about all these things that raises so many concerns down to just what is a relationship and is it valuable? And on some level, as a society, we have believed for centuries that a relationship is a healing space. There's a reason we pay therapists and we don't just pay to put.

    People in a room with exercise equipment by themselves like, like we have really valued a relationship as a healing space. And what is this new kind of like pseudo relationship based on what we've experienced, again, in social media pseudo relationships. Can lead to all kinds of trouble. And it just makes me concerned to think about more pseudo relationships happening when we think of ethical violations.

    Therapists commit these. All the time as well. Like we do not perfectly meet our code of ethics. We forget to do our continuing education. We, like I've mentioned at the top, like we're constantly getting our licenses suspended, revoked and we've decided as a society we need to be watching for therapists and all healthcare providers and really monitoring what they're doing.

    But how are. Those ethical concerns different than what's going on with chatbots?

  44. Megan Cornish

    First of all, there is a very strong mechanism we talked about earlier, to monitor therapists. Not only does that mechanism exist, we talked about licensing boards and reports, and, but people know about it, which, it's two separate heavy lifts.

    First of all, we don't have the regulatory systems and the oversight. Now the only recourse people have is lawsuits, which is a big deal. But no one's gonna be able to call a hotline like we do with therapists to report a chat bot, also a therapist, over the course of my career, how many?

    Maybe a thousand clients, maybe more. But that's the exposure risk for me. Whereas a AI chat bot is going to see literally millions of people a day so that the potential for harm is much greater and the recourse for those who are harmed is much smaller. And we, I can lose my license. You could lose your license.

    That's a real risk. But the amount of power with all the money that they have, these companies are, it's gonna be a PR issue. They'll get a fine slapped on them. And, but there's no world where they, lose their ability to be a business. And of course there's also things like consent that we need to talk about because when a client enters into therapeutic or relationship with us we're very careful around what they.

    Them understanding what they're doing and being affirmatively, like saying, yes, I want this, and there's no such requirements on the part of these AI companies, as well as I'm convinced that even, I work with these companies that even I don't know what they would be consenting to because there's so much under the hood that we don't know.

    So I think that's another part of the difference between therapists and ai.

  45. Sarah Lyon

    There's so many like weighty things. As we start to talk about them, it really starts to feel. Overwhelming and that we're up against Goliath. I was just thinking about your point about lawsuits. I was like, if you sue our local therapy clinic down the streets, it's a pretty even fight. Like they are not the most resourced person.

    There's individuals in my town that are more resourced than our local therapy clinic. If I go to Sue Google, these huge like open ai that is not an even battle. Like it just the scale or the uneven of it, unevenness of it is so stark. Like you literally can't even visualize like

  46. Megan Cornish

    Yeah,

  47. Sarah Lyon

    the billions versus are resourced hundreds of thousands.

    Like it's just, yeah, it's so overwhelming, which leads me to be like as individuals in this. Incredibly tilted battle. What do we do? What's our ethical duties? We're still under our code of ethics. We're still under our license. What do we do as individual therapists? I think especially just thinking about like session to session, what do we do with this huge moral quagmire hanging out there?

  48. Megan Cornish

    Yeah, I think that's a really important question. You read the APAs like statement of ethics. Social workers actually have their own ethical code, and one of the things in our code of ethics is that, is advocacy that we need to be actively advocating. I would love to see every therapy profession adopt that as their code of ethics because we need to be advocating for regulation right now.

    I think states are the ones who are trying to step in the gap right now because we're not really expecting a whole lot from the federal government at this moment. And the federal government is actually trying to tamp down on the state's regulations of ai. But we need to be out there encouraging our state lawmakers and our representatives in the federal government as well that something is better than nothing.

    Some of the laws that are coming, ours get, they're getting a lot of pushback. Especially from the business community, of course, that they, of all the reasons why they're not perfect or they're gonna, make something worse, but in my opinion, something is better than nothing. We can always come back and revise, but if we have to wait for the regulations to be perfect, then we're gonna be waiting a long time and we'll be way past AI by that point.

    I also think it's the responsibility of therapists to be literate around ai. That they don't have to go get their, degree in computer science, but they need to have a basic understanding of the mechanisms of what AI does and as well as, what's the incentives in terms of who's funding these and how do they make money Just to keep an eye on the products that are being sold to you, as well as being sold to your clients and your, the health systems that you're a part of.

    I think we also need to stop seeding the conversation in public to tech therapists need to be actively participating in these conversations and not from a, our hair's on fire. We have to stop this right now position, but from a. We all have our licenses, we all have education. We all know what is important in therapy, and it's not that hard to see what's missing in the tools.

    So to be active about that, to make sure that we are the ones who are still retaining our clinical credibility and not letting tech grows who, have never taken care of a patient. Pretend like they know what they're talking about when we're talking about patient care. And I also think it's our responsibility to talk to clients about their AI use.

    And to actively ask, if they're using ai, what kind of advice they're getting from ai, how they decide whether or not to take that advice from AI when it comes to like more directive. Care plans. Sometimes therapists and mental health have like an open ended, like we're gonna see what happens.

    We're gonna try to do a few different things. But sometimes they have very measured direct approach where, first we're gonna do a, then we're gonna do B. And I think that's probably how therapists are physical therapists. Occupational therapists are more like two. So being clear early on that if. If they're using ai, you don't judge them because they're probably gonna use ai, but before they make any decisions about what they're going to do, you'd like to hear about it because you're the therapist that's in charge of their care and you don't want them to start doing anything that's going through a wrench in the approach that you're taking.

  49. Sarah Lyon

    I wanna pivot to ask a couple more questions about our national response, but linger on the individual one for like our individual duty for. Moment. Should we be asking all of our patients, especially outpatient, about how they're using ai? Is that just part of our,

  50. Megan Cornish

    that's kind, we should be doing what?

    Our conversation going to mental health. So obviously I don't know how active your professional organizations are, but in mental health the advice right now is yes, you should be talking to your pa, your clients about AI use. And especially with the risk of AI, psychosis and mental health like measuring their understanding of how accurate AI is and making sure they understand the propensity of AI to hallucinate the synco of ai.

    All those things are really important to. To be protective of their mental health and their experiences with ai. So I, I speak from a place of mental health, but we are saying you should have a conversation with your clients about ai.

  51. Sarah Lyon

    What's like the gentlest way to ask about AI usage or have you heard recommended language about that?

  52. Megan Cornish

    You know what, I haven't heard recommended language, mostly because therapists don't tend to be shy about bringing up hard conversation. That's our job. Yeah. I think. I think that if you do it in a nonjudgmental way, I don't see it being an issue. Just putting the question out there, and especially if you're not coming about it as a way of telling them not to do it.

    Just asking like, Hey, do you ever look any of the stuff up on ai? And they'll tell you whether they do, and. Using that as a, just opening to, to hear how they take that advice, if they are actually running with it, if they're just getting ideas for what to talk to you about.

    I think there's, there's no reason it has to be a confrontational conversation.

  53. Sarah Lyon

    Thinking on. Like a broader scale, like a national scale. I have this like working theory about therapists that we have just been a sleeping voice. Like we one, we're really silo, we're really siloed from each other, like pTs and oTs and mental health. Like we don't even think of ourselves in the same bucket as therapists, but together in our country, there's around 2 million of us. Like we are. There's a lot of us. I know we're all we play important roles in our communities and I feel like we've underutilized our voice.

    Do you agree with that? And what are ways we could do better?

  54. Megan Cornish

    Yeah, I definitely agree with that. I think I see the attempts by individual therapists or even our governing bodies to get us all on the same page, and there's sadly just like a lot of mistrust that prevents us from doing that. But I do think that, a group of 2 million people.

    Around any issue would be hugely influential.

    The big problem being like, who's gonna lead it? Who's gonna be trustworthy enough to lead it? So if, if I had a magic wand and could figure out that problem, I think a lot of other things would fall into place. But I don't have,

  55. Sarah Lyon

    yes. Yep.

    We do have the power, we do have the voice. We just need to figure out

  56. Megan Cornish

    how

  57. Sarah Lyon

    to wield it, and maybe it will take this huge external problem. Like this. Yeah,

  58. Megan Cornish

    And I think part of the problem is we, social media hasn't helped, but we've all become so individualistic that we, that the idea of not being the leader and instead just being a supporter, like somehow feels distasteful to us.

    And that's more of a like individualistic cultural problem. There are other cultures that don't have that same problem. So learning how to like value, just throwing your small voice behind a larger cause and like seeing that as like a worthwhile use of our time as humans, I think would probably be the first step.

  59. Sarah Lyon

    And if anyone can figure it out, I truly believe it's therapists.

  60. Megan Cornish

    Yes, exactly.

  61. Sarah Lyon

    Therapists are awesome.

  62. Megan Cornish

    Yes.

  63. Sarah Lyon

    We began. With kind of your mood around this subject, as we head into our final five minutes, I wanna ask two questions. One, what's your big existential dread around this? Let's just say the worst, let's say the dread out loud.

    And then what's your big existential hope? Like where do we hope this could all go? Let's start with the dread and end on

  64. Megan Cornish

    the hope. Yes. The dread is easy because the current trajectory seems to be aimed right. In the direction of the dread, which is we're going to do nothing as large tech companies extract all resources from the environment and the individuals in order to continue to have all the wealth and all the power and leave nothing for anyone else and nothing for the environment in the interest of sitting on big piles of money.

    So obviously that is the dread and that does seem to be how we're starting now.

  65. Sarah Lyon

    Yeah,

  66. Megan Cornish

    but I also I recently read Ezra Klein's book Abundance, and he said something in the intro that just has stuck with me. He articulated it so well. I know we talk about universal BA basic income, and even, some of these billionaires are paying lip service too.

    UBI. But Ezra Klein said, since we, since AI was built, using the collective wisdom of all of humanity, its resources are shared out equally among them. And I'm like, oh, that makes total sense because all the information that went in to make ai, what it is came from all of us and from our ancestors and everything that we've compiled in the form of the internet has come from all of humanity.

    So why then would we not. Also equally benefit. And right now the race of AI is to make better ai. That's where all the resources. But if it were to change, to solve even the only the issues itself, AI itself is causing like it's use of water and issues on, around the environment and poverty and wealth inequity.

    If. We were to change the resources to focus on those pro solving those problems, I really feel like it would make a big difference to find the technology to solve all these huge issues that we're facing. But it would require those at the helm to decide that they, their pile of money is big enough and they're willing to sacrifice us, not having a second yacht to solve these problems.

  67. Sarah Lyon

    Yeah. I've reached a point where. I am just like. So many of our problems are now solv solvable. Like I think of that in healthcare a lot. I'm like, we know what good healthcare looks like. Yeah. We know what the delivery would look like. Maybe we can't solve every problem, but we can sure solve a whole lot of them.

    But it is going to take a collective will yes to the

  68. Megan Cornish

    other research is, but we lack the will.

  69. Sarah Lyon

    We lack the will and that, I think it's just, I see that as a call to therapists to use our skillset in a way that we maybe haven't thought of to whatever it looks like to rally our com, our communities, and really be talking about what is our will with these tools that we have?

    Yeah.

  70. Megan Cornish

    Yeah, I would love to see everyone talking about it so much that it becomes so normalized that everyone's like. That it's weird that they're not using it. Where everyone's like, why aren't they, this is strange. Right now. It seems like talking about it is a strange thing.

    Talking about it in a different way is a strange thing. But if people start shifting the conversation enough, like why couldn't we change the narrative where it seems like the strange thing is how they're using AI right now.

  71. Sarah Lyon

    Yeah. And I really think that's as simple as it can just be talking with your friends.

    It can be talking with your family, it can be talking with your organization, it can be talking nationally, like wherever you feel comfortable. All of those conversations are important and no one is more equipped to have them than therapists. Yes. And we have to be leveraging that skillset in this way.

    I had one, question come in at the end that I want to read to you. Alana said, I work with adolescents and see a big opportunities for OTs to educate on how to safely and effectively utilize tech for health management. Do you have any go-to resources on client education for AI literacy. Any tips for any of us on AI literacy?

  72. Megan Cornish

    So I actually have spent a lot of time in the past few weeks, I wrote a Substack article about teaching kids about ai. I had experience where I was like, I'm gonna be parent of the year. I've been talking about AI to my kids for a few weeks now. Now I'm gonna let it have a convers them have a conversation with ai.

    And I gave the instructions to AI where I was like, okay, I'm teaching my kids about ai, so I want you to emphasize that you are not a person. And they asked, what's your favorite color? I don't have a favorite color. What's your name? I don't have a name. It went really great. I ended the conversation and I said to my kids, how, what do you guys think about that?

    And my five-year-old said, I love her. She's my best friend. I'm like, cool. So my, my back petting went away. So

  73. Sarah Lyon

    I ended up going into all the research with like, how do I do this? And they answer is not to,

  74. Megan Cornish

    my first instinct is just to throw away all the AI and not let them have any use of it until they're 18.

    But that's the wrong answer too. There's not a ton of good research right now. I'm sure that's going to come right now. My desire is for them to have such good, robust width of experience with real relationships that when, by the time they have access to AI on a daily basis, that they will know that it's not the real thing.

    The analogy I'm using is it's like a apple pie scented candle. I want 'em to know what real apple pie is so well that when a something pops up on their phone and says, would you like the scented candle? It's exactly as good as apple pie. They'll be like, no. It isn't. It's just, it's not real apple pie.

    It just smells like apple pie. And they'll know that there's a time and place for a scented candle, but it's never going to be apple pie. So that's I'm proactively going to be teaching them about what relationships are and what the benefits of the friction is, so that by the time they're ready to like actually have usage around ai, that they'll know the difference.

    But my kids are young, so that certainly helps.

  75. Sarah Lyon

    And as you say that, I'm like, part of our skill as therapists is just giving language to the unfamiliar and hard things. And the more we can get language around this, like deceptive empathy, like lack of contextual awareness and just get that language out, whether to our kids or whether to tech companies or our national organizations, just all the better.

    We've talked about so many important things. I led this by being like, this is the conversation right now. What's the final thought you wanna leave us on? We've talked about so many important things. I know there's so many to choose from.

  76. Megan Cornish

    Ooh, my final thought. My final thought is there has never been a great cultural change without crisis.

    And I suspect that we are at a reaching or at a time of crisis with the, all the things that are happening at the same time right now and are. Our country and with technology and for those of us who have dedicated our life to helping people, it's a really good time to start taking more of a leadership role in our communities because we are going to need each other and we're going to need leaders who are courageous and hopeful to take advantage of the moment of crisis, to lead, to change and greater interdependence and less dependence on, what we can buy and more dependence on the lives that we can make for ourselves and our children.

  77. Sarah Lyon

    It truly is a time for moral courage, which I know therapists have and are oftentimes just uniquely gifted at because of our work. I had been looking for someone to have this difficult conversation on the podcast, Megan, and I'm so thankful that you said yes, and we're willing to just dive into this.

    Expansive. Tough. Ever moving topic. I so appreciate your time today. I found it so helpful, and I know I'm going to be chewing on it for a while, so thank you for the time today.

  78. Megan Cornish

    Thank you for having me.

  79. Sarah Lyon

    Thank you for joining us on the OT Potential podcast to earn one hour of a OTA approved continuing education for your time today. You will need to sign in or sign up at otpotential.com. Once you're in the OT Potential Club, you will find a five question post-course quiz connected to this episode. When you pass the quiz with a score of 75% or higher, you will be able to download a PDF certificate that certifies your completion of this.

    Okay. I wanna thank you for joining us today and we'll see you next time.

Professions

AOTA-approved for OT CE in all 50 states

Also relevant for these professions. Check your state's CE requirements before claiming credit.

Contact Hours (CEUs/PDUs)

1 Contact Hour. Released online from Aurora, NE.

This course was designed to meet your continuing education requirements

We designed our courses to meet the requirements for “online” and “independent/self-study” courses. To verify the requirements from your specific state (within the U.S.), check out OT Continuing Education Requirements. If you are outside of the United States and don't see requirements for your specific location on that page, please contact us.

OT Potential courses are presented as live webinars, then released on-demand video and audio via the OT Potential Podcast. Attendance in the live webinar will be indicated on your certificate. Learn more about how our courses help meet live webinar requirements, and OT CEU requirements overall.

We are proud to be an AOTA Approved Provider and to meet the requirements for your NBCOT renewal.

OT Potential Reports CE/CME to CE Broker in 24 Hours NBCOT Logo
AOTA Approved Provider Badge with Course Details
AOTA Details

AOTA Course ID#: 15909
AOTA CEUs Earned: 0.1
AOTA Delivery Format: Distance Learning-Independent
AOTA Category: OT Service Delivery
AOTA Level: Introductory

NBCOT Logo
NBCOT Details

NBCOT PDUs Earned: 1.25

Instructional Methods, Registration, Special Needs Requests, Cancellation Policy

This course is an independent/self-study course delivered via podcast on one or more of the following platforms: YouTube, Apple Podcasts, Spotify, & Google Play.

If you need accommodations to take this course, please contact us and we will address your needs on an individual basis.

If we cancel a promoted course, event, live stream, or any other paid CEU offering prior to release, and you subscribe explicitly for said offering, you are eligible for a full refund if you did not complete and earn any other CEU quizzes or certificates during your subscription.

If a live webinar is cancelled that you signed up for, our system will automatically generate an email to you and if possible, we will inform you of the rescheduled date. Our most current webinar schedule will be found here.

Audience / Education Level & Course Completion Requirements

Our target audience is therapy professionals who are looking to learn more about this topic. The educational level is introductory. In order to receive a certificate for this course, you must first participate in the podcast/webinar in its entirety. Then, you will need to take the quiz that will accompany the course and earn 75% or higher. If you pass, a PDF certificate will be made available to you via email and download. Quizzes for live (distance learning–interactive) webinars must be completed within 3 days of completing the webinar.

Disclosures

It is the policy of OT Potential to disclose any financial and non-financial interest the provider or instructor may have in a product or service mentioned during an activity. This is to ensure that the audience is made aware of any bias of the speaker.

OT Potential pays guest speakers an honorarium for their participation, either directly or to an institution of the speaker's choosing at their request. We invite guests whose expertise often comes from practices, programs, or building products in their field. Where a speaker has a financial interest in the subject of a course, we disclose it below so you can weigh it for yourself.

OT Potential's interest in this topic: We have no financial stake in this topic.

Speaker interests: Megan Cornish runs Megan Cornish Marketing, a strategic communication and marketing consultancy serving mental health technology companies, and contracts with Allia Health as Head of Provider Growth. She has a financial interest in the mental health technology sector discussed in this course.

Speakers

Megan Cornish

LICSW

Speaker

Megan Cornish (LICSW) is a consultant and strategist who works with behavioral health startups on product-market fit, go-to-market strategy, and messaging. A licensed clinical social worker by training, she left direct practice to focus on the business side of mental healthcare, and found herself helping companies bridge the gap between what therapists (and clients) need and what the market builds. She writes Incomes & Outcomes on Substack and won a 2025 Best Story Award from Psychotherapy Networker. She lives in Richmond, Virginia with her husband and three kids.

Sarah Lyon

OTR/L

Speaker

Sarah Lyon, OTR/L, is the CEO of OT Potential. Sarah earned her BA from St. Olaf College and her master’s degree in occupational therapy from New York University. Her diverse clinical background spans multiple settings, including critical access, acute trauma, and state inpatient psychiatric hospitals. In 2011, she founded OT Potential to fulfill the industry's need for reliable, high-quality occupational therapy resources and continuing education.

As a recognized content creator, Sarah has collaborated with top healthcare brands like VeryWell Health, WebPT, and MedBridge. She blends her clinical expertise with a talent for creating clear, action-oriented content that empowers practitioners to excel. Passionate about elevating the OT profession, she has been featured on numerous industry podcasts. Sarah ultimately returned to her roots, running OT Potential and raising her family in her hometown of Aurora, Nebraska.