AI is changing therapy with Dr Aisha Tariq

Welcome to the Business of Psychology podcast. Almost every psychologist and therapist I speak to right now has some version of the same feeling about AI. A mix of curiosity, a strange sense of guilt, and a worry that they're either being left behind or they're about to do something unethical without realizing it. There is a lot of noise, a lot of hype, and very little practical talk about what actually helps in a real practice, and where the genuine risks of AI might be. My guest today has been doing AI since 2022, long before most of this conversation started. Dr Aisha Tariq is a clinical psychologist and the clinical director of Illuminated Thinking, a Glasgow practice of 25 clinicians. She's built AI into how that practice runs and how her team work clinically, so she can be really specific about what saves real time across her team, and also where the pitfalls are. She's also built her own AI clinical documentation tool, Cogent, and she's the co-founder of the AI in Psychology Summit, which is happening on the 5th of October in 2026.

Full show notes and a transcript of this episode are available at The Business of Psychology

Links:

AI in Psychology Summit - 5th October

Esther and Aisha are offering listeners 20% off the summit tickets!

Follow this link for the discount: https://aipsychology.co.uk/?ref=DRROSIE&discount=PODCAST20

Episode 162: Using AI in therapy and other legal questions with Clare Veal from Aubergine Legal

Links for Aisha:

Cogent Clinic

Links for Rosie:

Substack: substack.com/@drrosie

Rosie on Instagram:

@rosiegilderthorp

@drrosiegilderthorp

The highlights

• Aisha tells us about her independent practice and she ended up with a practice of 25 clinicians 02:23
• I ask Aisha what it was that made her take an interest in AI 04:22
• We discuss the feeling of shame associated with using AI 10:29
• Aisha talks about rolling out AI to her team 13:42
• Aisha tells us what steps to take if we'd like to explore having these tools in our practices19:50
• Aisha talks about the development of Cogent 24:13
• I ask Aisha what common mistakes she sees psychologists and therapists making with AI 32:35
• We discuss the AI in Psychology Summit 33:54

Start Up Your Psychology Practice

If you're fitting private practice clients around an employed role that doesn't give you the flexibility that you need, this is for you.

Start Up Your Psychology Practice helps you build a practice that can replace your income with fees that work, a clear specialism and marketing that fits your daily life so you can finally let go of the juggle.

Right now it includes a two-hour one-to-one intensive with me to plan it all out together. Follow the link to find out more and book a call with me to see if Start Up might be right for you: https://psychologybusinessschool.com/lp/start-up-private-practice-bundle/

Speakers:

Rosie Gilderthorp

Aisha Tariq

Welcome to the Business of Psychology podcast. Almost every psychologist and therapist I speak to right now has some version of the same feeling about AI. A mix of curiosity, a strange sense of guilt, and a worry that they're either being left behind or they're about to do something unethical without realizing it. There is a lot of noise, a lot of hype, and very little practical talk about what actually helps in a real practice, and where the genuine risks of AI might be. My guest today has been doing AI since 2022, long before most of this conversation started. Dr Aisha Tariq is a clinical psychologist and the clinical director of Illuminated Thinking, a Glasgow practice of 25 clinicians. She's built AI into how that practice runs and how her team work clinically, so she can be really specific about what saves real time across her team, and also where the pitfalls are. She's also built her own AI clinical documentation tool, Cogent, and she's the co-founder of the AI in Psychology Summit, which is happening on the 5th of October in 2026. So before we start, a declaration from both of us. Aisha has a commercial interest in this area because of Cogent, which we will be talking about, and I have an affiliate arrangement with the summit, so I earn a percentage of ticket sales through my link. We're not gonna spend this episode selling you stuff. This is all about the wider picture, because what is useful here isn't about the products, it's really about the concepts and how we use AI going forward in this kind of inevitable feeling industrial revolution that we're in.

So welcome to the podcast, Aisha. It's really great to have you here today. Could you start by just telling us a bit about your independent practice and how you ended up with a practice of 25 clinicians?

Yes, of course. Thank you so much for having me on, Rosie. So I started in private practice in 2019, and I think the same reason that most of us go into private practice, which is for a better work-life balance. At that point, I had just had my third daughter. She was two at the time, and I just... yeah, I needed the time and space to be with my family and also work. And so I started my private practice in 2019, and in 2020 I decided to leave the NHS. And from there, I guess, things grew rather than being specifically planned. I think that the demand that was coming in at the time, I thought it'd be a good idea to take on some associates to help meet that demand locally here in Glasgow. And before I knew it, I had 20-plus clinicians, five therapy rooms, and staff that I was managing.

Wow. So it kinda turned into a big job quite quickly.

Yes, it did.

So just, 'cause I know a lot of people will be wondering, you know, how that happened-

have you got a particular clinical specialty that you work with?

My specialism is in trauma. That's what I've always worked in. So when I was in the NHS, I qualified in 2014, and I'd always been interested in trauma and approaches to that. And one of the big things I had wanted to do when I left the NHS was train in the kinds of therapies that I wanted to rather than waiting on funding coming in. The very, very first thing I did when I left was my EMDR training, and it was the best thing I could have done. And then from then, I've done IFS and schema and various things. But that's one of the things I've loved about private practice, is being able to choose your specialism, choose what you want to develop in and how as a clinician to. So the majority of people that come to us are for trauma.

And I guess suddenly having that many people to manage, that's a lot of clients coming through, but it's a lot of clinicians to, to kind of balance the needs of as well. I'm wondering if that maybe made AI seem quite attractive to you. What, what was it that made you kind of take an interest in AI?

It was definitely one of the big reasons, Rosie. So, a couple of years ago... Well, it was at the... It was 2022, I think it was, when AI kind of really started to take hold, in the, in the, in the field, I guess, in IT and in infrastructure. It became kind of the possibility started to open up around that time, and it was around that time I have this vivid memory of sitting at 11:00 PM in front of the TV with my laptop open, sitting next to my husband. We're trying to watch something, but I'm also trying to do some admin, and he commented and he said, "You're working more than you did when you were in the NHS. At least that was a 9:00 to 5:00 job." And it was a kind of moment of, ah, yeah, you're right. What am I doing? This isn't what I came into private practice for. And so I decided to use this hobby of developing tools using coding and my understanding, very basic understanding of AI, and try and bring that into the practice a little bit. I had always been the geek at uni who liked SPSS. I don't know if you remember that and the tools?

Oh my God, I'll never forget SPSS. But I am amazed that you liked it.

Oh, I was the one, I was the one who'd be sitting there geeking out in the library on all these tools, and what they could do and the possibilities, and I was always amazed at what the technology could do and how we were always using it. There was always such a disparity. And so this was, this was another moment where I thought, "Actually, can we use this in a way that we're not using it just now?" It was a wee bit of... Well, actually it was a huge amount of trial and error. I very rarely come across anyone in private practice or generally within mental health who's using the AI and technology aspect of things in the way that I am, and so sometimes it can feel quite isolating, a little bit of shooting in the dark a wee bit with it, but yes, that's where my interest came from.

I mean, I just think that's a really great way of thinking about AI. I think unfortunately, because of the products that came to market first for the general public, a lot of people are using AI to do things that actually humans are better at, like creative writing. You know, I sometimes use AI to help me structure an email that I've written because, this will not surprise any listener, I can be a bit flow of consciousness sometimes . And AI sometimes can pick up on that and say, "Hey, a subheading here would be useful." So it does do some, you know, useful content generation stuff. But generally, I think this, this kind of productivity aspect of AI, where can we use AI to do the stuff where humans make a lot of error and take a lot of time to do something that is actually very simple? I think that's just the healthiest human way of using AI. And it gives us more time to think.

Yes, absolutely. I actually think we've came to AI from the wrong angle. I think we all started with clinical notes, but that's the bit that we give a part of ourselves to. The only bit that's left of us in a clinical record, in a clinical note. I actually think the repetitive, mundane admin tasks of a practice are where we should be focusing our time and allowing AI to take over the bits that clinicians don't like doing. You know, I think that's where we should be using it, rather than the bits that require judgment in our clinical knowledge and skill.

I completely agree. And I've got this theory that we might be, we might have been for the past 10 years, in, like, the really rubbish part of the technology revolution, where we have all this tech, but it actually just gets in the way of living as a human, and so you're doing more admin because the technology has made it possible to do more and more admin. And I've been very jealous of my previous self in my first few jobs as a psychology assistant, where all I had was a notebook.

Yeah.

And my notes would be in that notebook, and no one expected them to be anywhere else. They were just in the notebook. And that was okay. And I just think we've created, with technology, more and more admin, and now is our opportunity for AI to take some of that admin away again and get us closer to living in the real world and actually thinking about other people. And you know that awful thing where you go to an appointment, whether it's... I mean, it hasn't happened to me in a therapy appointment, in fairness, but I'm sure it does sometimes. But you go to your GP or somewhere, they're just typing the entire way through talking to you. And you're like, "I could have a giant blue pustule on my neck and you wouldn't notice, because you're not looking at me. You're just typing."

Yes.

And I love the idea that AI can change some of that, the kind of boring bit that stops us from connecting.

Yes, absolutely. 100% agree with you. And Rosie, the thing is that in other areas, in other fields, people are using AI in that way. I feel like in some ways psychology and mental health has been overly cautious, understandably cautious, but perhaps overly so in adopting AI. I think the conversations I have with clinicians are often, "I kind of want to use it, but I'm worried about if I'm going to use it rightly or wrongly." I think that's such a big, a big thing. And then for the clinicians who are using it, there's often a lot of kind of quiet shame around using it. You know, people don't admit that they use AI for things. And so I think that that's, that's a big part of what needs to happen next is these open conversations. I think the more quieter we are, the more likely we are to use it badly. You know, if we use it openly, I think we're, we're gonna compare notes and be able to make better use of AI.

Absolutely. I mean, what do you think that shame is about? 'Cause I've really noticed it too, that people don't really like talking about it. It feels quite exposing when you say in a room full of your peers, "I'm trying this AI tool out," 'cause you know 50% are going to be judging.

Yes.

But what is that?

I think that, I have a lot of different theories about it, and there's probably lots of different reasons for it as well. I think one of the, the big things is there are no guidelines. There is no rule book. There are no boundaries within psychology. No one's told us what's okay and what's not okay. And so there's a bit of kind of putting yourself out there with something that, you know, maybe, yeah, we just don't have the answers to in terms of how we use it. No one's told us. The BPS, the HCPC, you know, they're still developing these guidelines, and so it's still such early days. I think the other part of it is the controversial side of AI. You know, the horror stories we hear in the news of how AI can make situations worse or, you know, very dangerous, and used inappropriately. And then there's the other part of it where there's this kind of implication that AI is doing your job for you So you're giving something so that you don't have to do it. You know, I don't know if that's part of it too. I think there's lots of different aspects of that, of that shame.

Yeah. It's really interesting, and I think all of us will remember that online therapy was the same.

Yeah.

And, I started doing online therapy in, I think 2017, maybe 2016, and everyone was so anxious about it. I did a whole certification in safe online therapy, which was really intense. It was like another therapy training.

Oh, wow.

And I remember throughout the whole thing thinking, "I don't need to do this. I don't need to do this." Like, so long as you've got your kind of ethical practice code, you actually... I don't think that was necessary. But there was that much anxiety about it that I paid a lot of money and did a lot of work just to kind of deal with that within myself.

Yeah.

Rosie Gilderthorp:

Yes.

And so I wonder what that moment will be with AI, and, and maybe we're kind of in the middle of it actually where it's becoming such a big part of every other aspect of people's lives that there'll come a point where if you're not using it, you're kind of not doing what your clients expect anymore, or you won't be productive at the level that they're expecting you to be. But I guess that's a conversation we could kind of get into for hours. What happened next in your journey? So what was the first thing that you rolled out to your team using AI?

So I had been using, um... So I built a few tools before AI, so my team were kind of used to me putting things in the practice, putting things in place that maybe were a little bit experimental. Just things around invoicing and the admin side of things, managing referrals in a more automated way. The initial kind of AI tools that were becoming commercially available, they weren't necessarily building stuff for us. So back in 2022, what ChatGPT could do was examine a piece of code that I had already written and say whether or not it was gonna do what I wanted it to do. So I guess that's where it all started, and I found it kind of very good for that kind of thing. And then it grew into actually could write the code as well. And that was kind of around 2023, it started doing that really, really well. And then from there it, it just has developed hugely. And now I know of clinicians who ship their own tools to production level, you know, through solely with AI use. And so there's been this kind of development of what AI can do. How I use it currently is I don't use AI as a chatbot anymore to have conversations with. There's this kind of development in AI called agentic AI, which is autonomous AI agents doing the work for you. So that's how we use it in our practice currently. To do the things that I used to do that I would end up doing every day or every week that were very monotonous, very repetitive, very boring. The things that I don't like doing as a clinician, that's what my AI use is currently.

Ah, interesting. So with an AI agent, you train it to do a task and then it just does it and reports to you when it's done it?

Yes. Yes, exactly that.

Love that.

So for example, one of my AI agents is a referrals manager. So what it will do is monitor referrals coming into the practice through all the different streams that we have, and it comes into a centralized dashboard. It will give me suggestions on which clinician would best fit for that referral based on their current availability, because it'll look through the diary. And then it-

My God, I love that. I think everyone with a group practice now is getting very excited.

And then what that AI agent does is send me a WhatsApp message to say that, "These referrals have came in today. These are the clinicians. What do you think?" And then I'll do the kind of checking over, managing any expectations, and doing the communication part, I'll do that bit myself. But even just that has taken off such a big, huge workload, the things I used to... You know, there's so much juggling of diaries, and managing who has availability when, what kind of presentation this is, and matching with a clinician's interest. And AI can do that very, very well.

And it's a brilliant example of AI doing something better than we can do it, because with as many clinicians as you have, you couldn't possibly hold all their availability in your mind while reading a referral. That would be impossible for our capacity. And I think, you know, inevitably, that's how we end up always going to the same few people. It's not our fault, we can't help it, but we are human and AI just doesn't have that barrier, does it?

It really doesn't. It could hold a huge amount of data at any one time, and make decisions based on that. And we have to have... you know, we have put a lot of guardrails in place around, you know, for example, risk. You know, if risk is flagged, immediately it comes to me, you know, before it does any triaging. There's things around specific conditions that, you know, I know that maybe need greater care. That will come through to me directly. Lots of things that we need to put in place before we can just go in and use AI. But yes, it's transformed our practice for the better. What I will say is we can't use tools commercially, you know, in terms of what's available to us as customers. We do need to be careful about what AI we're using, what AI is touching our clinical data. That's incredibly important, so I would just put that caveat in there as well.

Yes, because, you know, as you were talking, I was thinking, I think something like Claude has capabilities where you'd be able to build something like that using their infrastructure. But- I mean, we were talking before we started recording about how I don't trust any of them. And I don't think it has the high enough level of protection for health data. And, and that's why there is this distinction. So we might come into contact with other businesses who, we might have an interaction with for something other than healthcare, maybe, you know, shopping or e-commerce, something like that, and they might be using all of this stuff, and we can see them using that, and, you know, we might know that it's been built with something like Claude or ChatGPT. But we can't, we can't go off and do that because we have to have this higher level of data protection because the data we collect is so sensitive.

Yes, absolutely. So I would caution very strongly against using any of the commercial tools for clinical information, even a first name, you know, because it's the first name plus the context that you're giving that AI tool, that could potentially make that data identifiable. So I'd be very, very careful about that. What we've done as a practice is we've invested in some infrastructure so that we can host local tools. So what I mean by that is, our AI agents sit, it sounds like they're talking about people, but they sit in our physical clinical practice on a server, and-

Oh, wow.

They don't touch, they don't touch the internet at all unless we have very specific guardrails. So for example, they only touch our email system for a specific task, like pulling an email. They're not affiliated or connected with ChatGPT or Claude or anything like that. These AI agents are, we own them as part of the practice. So that's how we've set it up to make sure that those, those safety, I guess GDPR compliance is in place.

Yeah, I think that's amazing, and it's another thing that you would've been able to do because you had this interest already. I think most people, I can imagine a lot of people listening are already sweating at the word server. I would have no idea how to put a server in my clinic or how that would work. So, you know, if somebody's listening to this and thinking, "Oh, I'd, I'd quite like to have some of this stuff, but I have no idea, like I'm not going to be able to set up my own server," what step could they take?

I'm using all these fancy terms. A server is just a hard drive. It's a fancy name for a hard drive that's connected to a private network. That's all that is. So, what I would say is doing some research into how other practices are using AI. I think that when we have these conversations, you know, and comparing notes, I think that's probably the best way of learning. I think we do it from each other. I think going outside of our field is really helpful for what AI can do, but it doesn't mean that that's what we should be using it for.

Yes, absolutely. So there is the difference... So there's the security side and making sure technically a product is compliant, as secure as it can be with the data that it's processing. And then there's the compliance side, so within the UK it's GDPR that we need to be compliant with, and the data that we have counts as special category data, so it's a kind of tier above kind of generic data that GDPR talks about. And so Heidi have made sure that they can process special category data. They have the compliance already set, and then the security side of things, that it's at a level that they feel comfortable that their data is secure and that their client's data is secure as well. So there is a higher benchmark for health data. And that's where Heidi comes in. So when I was talking earlier about general tools, I guess what I was meaning more was the consumer AI tools like ChatGPT and Claude and Perplexity and Groq, I guess those chatbots. Heidi's I guess a separate category, isn't it? Where healthcare data can be processed through their servers.

Yeah, thank you. That's really useful to understand. And, we did have an episode actually with Claire Veale, who I know is going to be speaking at the AI Summit, all about the kind of legal side of AI. So if you're interested in that, anybody, I will drop the the link in the show notes so you can go and geek out about the legal side, because it can sort of lower the anxiety level a bit to understand it. And I think that's why these conversations are so important, because if we don't sort of dive in and try and challenge ourselves to really understand this landscape, then it can just feel really frightening, and you feel like you're at the mercy of tech bosses who might have a really different value set. Whereas if you understand it in the way that you do, it's much, much more exciting. You, you feel more confident that you can navigate it.

I think it will take time. You know, I think our profession as a whole, it will take time for us to get to that stage where we feel confident, and it becomes part of our practices, just like online therapy has. I think you're right, Rosie, that we are on that journey with AI as well. We just need to get over these initial hurdles, don't we, around the risks and the ethical kind of issues around AI, and the security side of things. Absolutely.

Yes. And so I know that you kind of developed all this stuff for your own practice, and then started to wonder if this might be useful to other clinicians too, and developed a product called Cogent. Can you tell me a little bit about how that product has developed and what it does now?

Yes, absolutely. Cogent came from my frustration with the generic tools and the things that they... I mean, some of them are great, you know. Some of them do what they say on the tin, and they've been designed really well for that. But when I first started using Scribe, for example, I became more and more frustrated that we were kind of having to fit our practice to the tool rather than a tool being created for psychology specifically. Our work isn't just one client encounter, and then that client disappears, and then the next time they come in, it's as if it's a fresh appointment. You know, our work builds over time with the client, with their journey through therapy, and with the story, and as we get to know them, and our formulation is such a big part of the work that we do, and none of the, none of the tools there were fitting with that. So that's where Cogent came from, and I have been very fortunate that my poor associates have been willing to test things for me and allow them to be guinea pigs in all of this. So we've been using Cogent in some shape or form over the past two years within the practice, and I guess it was having more and more conversations with other practices, other clinicians who were interested in it and how I was using it. And so I thought rather than trying to- kind of piecemeal something together for specific people, I thought maybe this is something that there's a demand for more wider, and that's where Cogent came from.

I think there definitely will be a demand for it because from what I've heard, from what you've told me about it, I think it goes way beyond the kind of saving you a bit of time with your notes and into actually helping you to be a more clinically effective psychologist or therapist. So can you tell me a little bit about the stuff it does that isn't note-taking?

Yes, absolutely. I feel like the note-taking side of things is kind of like, yes, we all know, yeah, AI can do that, but I feel like there's so much more that AI can do specifically for our practices in mental health that nobody else is really doing. So one of the big parts of Cogent is helping with the other side of clinical practice, things like supervision, things like reflective practice, things like, you know, thinking through formulation and reformulation, how we think about a client over time. We've got functions in there that help us prepare for the next session. You know, sometimes you're scrambling looking back through clinical notes to say, "What was it I was working on here?" With Cogent, you click a button and it'll tell you exactly where you're at in that therapy journey and where you need to go next based on the notes that you've already created for that client. We have psychometric tools built in so you can send clients psychometric tools that they complete, and Cogent will auto-score for you. We also have a report writing tool. I feel like I'm listing lots of features to you, Rosie. So the report writing tool's been wonderful. We've had lots of feedback on it, and clinicians really valued it during the kind of testing phase of it. So it's ADHD and autism reports across the lifespan, and we also have more general report writing as well. We have a client-facing documentation there as well, and then specific therapy tools, and each of those can be adapted and be flexible to the clinician's preferences in terms of their writing style.

I really, I love all of that, and I think that there might be people that, you know, worry about report writing in particular, but actually, again, if we're honest about our own limitations, I think it... for me, I love writing, I really do, and I, really enjoy writing something for a client in the way that I think they will find it useful, but I would probably still use a tool like that just to make sure that I have encapsulated the whole of the history. Because when you've got a complex history, and I used to work in adult learning disability, so you often had a lot of life to work into an assessment and to try and make sure that you were really representing the wholeness of a person, and that's very difficult to do, and so I imagine a tool like this would be incredibly useful for pulling all of that together, and then you can still rewrite and put things into the words you think a client will, you know, relate to the best, that kind of human psychology element. But you can be more confident that you haven't missed something important from their, you know, third year of school that they told you about, but you've sort of, you know, got lost in the ether.

Yes. What I would say is, you know, I think when people think of AI tools, they think, like, this is a task that's going to be taken off my plate. I actually don't see it like that. I kind of see it as a very, very enthusiastic assistant that's very inexperienced, but really, really wants to help. And so everything that it's producing, you want to check it and make sure it's accurate, it's, relevant to what you're doing, and it's picking up on what you wanted it to do. So that's how I treat AI more generally. With Cogent, I was very, very careful that actually the thing that we do, which is the clinical judgment, our own opinion, our expertise, our skills, isn't taken away by AI. And so even, for example, the report writing tool, it's mandatory for you to put in your clinical opinion and what your findings are. It won't generate a report without that. Also things like, you know, observations that you made, your formulation, all these things are things you still have to type in to Cogent, that's not going to do it for you. I was really mindful that that was something that was important to me.

Yeah, I think that is really important, and will be... I think it's the difference between AI enhancing what we do, and the potential that I think is there for AI to kind of blandify everything to the point where it's not very useful. Like, I've definitely seen reports that I suspect have been written fully with AI and that I can't see the person in. And, you know, people have shown me examples of that to ask for my opinion, and I've been like, "I don't think that a properly qualified person wrote this, and I can't see, usually, your child in it at all." And that's horrendous. And I think probably some of that fear we were talking about earlier comes from people seeing things like that and thinking, "Oh my goodness, I would hate to put my name to that."

Yes, and I would very much term that as AI slop, Rosie-

Yeah, yeah ...

I don't think that's okay, and I don't think that's how clinicians should be using AI at all. I would say that that's where a lot gets defaulted to, to say, "Well, let's produce something." And, you know, AI's so good with language, I guess that's where it all started wasn't it, with a chatbot. It's just so confident with words. It makes you think what it's produced is good because you're like, "Oh, that sounds quite nice." But there's no, often no substance behind it. And so yeah, we need to be mindful of that.

Yeah, and I think that's why what you've done with Cogent is so helpful because if people can see this good example of how it can be used to help enhance your practice, give you more thinking space, more thinking time for each client, then I think we can start to discern the difference between a genuinely useful tool and one that is just really confident, and maybe affirms us in some ways but doesn't actually make our clinical practice any better.

Yes. One of the things that we need to do when we're developing a tool like this is put in place what they term as guardrails within AI. And so no one ever sees this part of Cogent, but we have hundreds and hundreds and hundreds of lines of guardrails behind the scenes. So every single button that's clicked, Cogent reads through these guardrails and makes sure that it's not producing anything that's outside of its competence, kind of like us as clinicians. So there's all these things that have to be in place that I think sometimes get missed when we're using generic tools. You know, a clinician won't know that we need to put certain guardrails in place when we're using a chatbot like ChatGPT, for example. So that's why, I guess, that's another caution against using those kinds of platforms.

Yeah, and I think we've probably touched on this already, but I just wanted to ask you, you know, what common mistakes you see psychologists and therapists making with AI? Are there any that we haven't covered yet?

I think the biggest mistake is probably thinking that AI is too risky, Rosie, you know? In a more kind of broader way, I think what I'm finding more and more, and I'm surprised often, is how much fear there is around AI use, and I think that's a very huge missed opportunity for a lot of people. I see, sometimes I have conversations with colleagues who have run really big practices, and they're run off the ground, you know, and often burnt out with how much admin they're doing, and it would be such an easy task for something like an AI tool to be able to take off of that person's plate. I think that's probably the thing that I see, that I feel kind of a bit sad about in our profession currently, you know, that there's so much potential there that's been wasted.

Yeah. I think it is a bit sad, and it reminds me of many other things which have been resisted in therapy, and then when, when the revolution eventually comes, everybody's very grateful that it did, but we do seem to be a profession that takes a while, a cautious while.

A cautious while. I guess it's one of the reasons why Dr Esther Cole and I wanted to have the summit, is to have those conversations, and to have those open conversations about how we're using AI, how we can use AI, and how we shouldn't use AI as well.

Yes, this is really exciting. So I'm gonna be attending the summit. I think it's the 5th of October, isn't it?

It is, yes.

So hopefully when this goes out, you've still got enough time to get yourself a ticket. I'd really recommend it. You've got an amazing line-up of speakers. Tell us a bit about who's gonna be there and what conversations you're gonna be opening up.

Yes. I'm so proud of, of the experts that we have on, on the day. I'm so pleased that we have such a breadth of experience. We have a big part of the day, I guess, is around ethics, ethical use of AI and considering the risks.  Dr Rima Lamba will be there talking about that, and Winnie who works within the mental health field and is an expert in using AI tools. So we've got a few sessions on that. We have Dr Natalie Stott, who I think is also a friend of the show coming on to talk about kinda everyday uses for AI, and the tools that we can use, and I think she's very much a proponent of we can go so far beyond just clinical note writing and a chatbot as well. So and then we've got Claire Veal, of course, coming to talk about the GDPR and compliance side. I know that that's a big worry and a concern for people. What I will say about that is often it's a fear of the unknown rather than it being about the actual AI itself. And so hopefully sessions like these will help clinicians feel more confident about where the line sits with our use of AI. We have Dr Naomi Murphy coming along as well, and she's going to close the day with a kind of reflective session on where the profession sits and where we are right now, where we need to go next with it all, and then we'll do a panel discussion of all of our speakers coming together at the end and answering questions through the day as well.

I think it's so exciting. I'm really looking forward to coming along and absorbing all of that and hearing all of those different points of view, because I fully agree with you. I think the more open conversation that we have, the more kind of non-judgmental as well, open conversations we can have, sort of accepting that this is a new frontier, nobody is doing it perfectly to start with. There will be things we try and then think, "Mm, actually I didn't like that, i'm not gonna do that anymore." We have to be open to that kind of learning process with this, and I think, the summit is providing a really nice container for those kind of maybe difficult conversations to happen.

I think it's important we have those difficult conversations. I'm hoping that people skeptical of AI will come along as well and bring their questions and concerns because I think it is so important that that is discussed. You know, I think we'll have people at both ends of the spectrum as well as the people somewhere in between, so I'm so excited to have those conversations.

Yeah, and I think just for anybody who might be feeling really anxious listening or feeling like, "Oh my goodness, this is so far outside of my expertise," I sometimes think it's really helpful to remember it's just another skill. Yes. And like all the other skills you've learnt as a, as a psychologist or a therapist, it does take time to get your head around it. You're not going to be an expert in this tomorrow. But actually, all we have to be is good enough at it. And so I think being willing to come along to something like the summit, immerse in it for a bit and give yourself that CPD opportunity is really valuable. And, just like just done a series on mindset blocks that often prevent people from really enjoying their practice or even starting it in the first place. One of the ones I always talk about is the belief that if I don't already know it, it's because there's something wrong with me. And never is that stronger than when it comes to tech. The number of people who at, like, 40 are like, "I'm too old, I can't understand it." I'm like, "Hold on, you are not too old. You're probably gonna be alive for another 50 years." We are never too old to understand this stuff. It's just that we won't make time for it, and I am really looking forward to making some proper time to understand as much as I need to about this stuff, and I'd encourage anybody else to do that.

Yeah. Thank you, Rosie. Yes, and we're looking forward to having you there and, and listeners there too. The other aspect of this summit that we've maybe not touched on so much is client use of AI. Rosie, that's such a big part of all of this too. You know, sometimes I, get bogged down and I love to geek out about how we can use AI in our practices, but the other part of it, perhaps even the scarier part of it, is how clients are using AI. And, you know, this is a big part of the conversations that I have with clinicians often is, you know, how to approach those conversations, what do we do when a client tells us that they're using AI? You know, where do we draw the line and the boundary to say that's okay versus that's not okay?

Yeah, and do we even have the right to do that? I think it's really tricky. I was really confronted by a client of mine recently who had pulled everything I've ever written, and there's so much of me online, so much. And there's, there's about to be even more when the book comes out. And, she basically said, "I want you to be Dr Rosie Gilderthorp, she's my therapist, be her."

Oh, wow.

And yeah. And, and in a way, it was really smart to do that, and I had to admire it, but I was also really freaked out, and I was most freaked out 'cause she showed me its responses and I was like, "That is what I would say, mostly, but not entirely."

Yeah.

And so yeah, it freaked me out, freaked me out a lot, and we had a lot of conversations about, you know, whether that was useful to her, and I tried to be completely non-judgmental about it, because I think there is an inevitability to that and that we, again, I've talked about this before on the podcast, you've always got to remember that there are people with a real vested interest in our clients using their chatbots in that way. And they have money that we'll never compete with. And so I think probably my stance on it is that it's better for us to work with clients about, you know, what's the beneficial way of using this, just like we would with anything else. You know, that's how I approach alcohol use, substances, everything. It's like, you know, what, what part of this is good for you and what part of it is not? And how do we find that kind of safe path with our clients? But yes, it's a whole learning curve. I can't wait for that discussion at the summit.

I completely agree with you, Rosie, about, you know, asking clients in the way that we would anything else as part of the kind of initial getting to know them stage. I think it's so important. I don't think clinicians really do ask, and I think a lot of clients are using AI. There was a study at the beginning of this year that identified a third of people accessing mental health services were also using AI chatbots.

Of course they are.

I think now it'll be even more than that with the kind of development of AI over the past year. And so I think we're missing a huge part of someone's life and how they understand themselves in the world if we don't ask.

Yeah. And I think we've also got to think about how we might have positive influence over how those chatbots are trained.

Yes. Absolutely. Yes.

A whole area of opportunity for psychologists and therapists, I think.

I think so, but I actually don't think it would take a huge amount of knowledge or skill about AI in order to do that, Rosie. I think, you know, for example, you know, one of the very, very first things that I did when I started asking clients was asking them what platform they were using and then handing them the prompt to use. Like, so say, for example, with ChatGPT, you know, I would type out a prompt for a particular client, you know, if they had OCD, for example, and I was concerned about reinforcing beliefs within, you know, their particular presentation, then I would kind of draft out a prompt for that person and put guardrails in place to say, "You're not to cross this line. You can't talk about this. You can't do this. This would be helpful, this would be unhelpful," and ask the client to start their conversations with that at the top. So things like that can be really helpful, and I don't think that that would require a huge amount of technical skill for clinicians to be able to do that. So these are the kinds of conversations we're hoping to have at the summit.

That's amazing. So many exciting things that we can start exploring, you know, together as a group of professionals. So the one thing I don't think you've told us yet actually is if people are interested in Cogent, where can they go and find out more about that?

Yes, absolutely. So, the website is cogent.clinic, but I will send you a link, Rosie, to the platform, and then hopefully you'll be able to use that in your show notes, if that's okay.

Amazing. Yes. I'll make sure that the link is there for you, 'cause I imagine lots of people will have heard about that and thought, "Hmm, that sounds perfect. I really want to go and explore it." So thank you so much for joining us today. Is there anything else that you wanted to mention before we get going?

No. Thank you so much for having me, Rosie. It's been an absolute pleasure. I've been a fan of the show for a very long time. I'm always listening in during my morning workouts. I work out in the morning, and it's lovely to start the day with your voice. So thank you for having me. It's been, it's been wonderful.

Aw, thank you so much for coming on.