AI in Design: Client Data and IP Policy's

Clients are starting to ask interior designers how they use AI with their private information — floor plans, budgets, family details, meeting transcripts. Rhiannon Lee, AI strategist for Australian interior designers, breaks down what actually happens to uploaded data, why your account type changes the answer, and how to build a simple studio policy so you can answer with confidence instead of stumbling through it.

Register for the free webinar (24 August):
→ What's really happening when you upload files to AI:

Explore Penelope, proposal GPT here

Get your Contract Clause HERE

──────────────────────────
CHAPTERS
──────────────────────────
Why clients are starting to ask about AI in interior design
What actually counts as "using AI" in your studio
Protecting your own IP from AI in client contracts
What client information should never go into AI
How to redact client details before you prompt AI
What actually happens when you delete an AI chat
Does your AI account type change your privacy protection
What is shadow AI and is it happening in your studio
Getting client consent to record and transcribe consults
How to build a simple AI policy for your studio

──────────────────────────
RESOURCES MENTIONED
──────────────────────────
→ Free webinar — What's really happening when you upload files to AI: [link]
→ DM "contract" on Instagram for the AI IP clause: @the_rhiannonlee
→ Studio Build — 6-week AI implementation intensive: [link]
→ Studio CEO — 12-week business coaching program: [link]
→ Studio Learn — free resources for interior designers: [link]

──────────────────────────
ABOUT RHIANNON LEE
──────────────────────────
AI strategist for Australian interior designers. Former Oleander & Finch. Creator of the Studio Suite — Studio Learn, Studio Build, and Studio CEO — operational AI implementation for design businesses, not productivity theatre.

→ Instagram: @the_rhiannonlee
→ Website: rhiannon-lee.com

Thanks for listening to this episode of "Designing Success: From Study to Studio"! Connect with me on social media for more business tips, and a real look behind the scenes of my own practicing design business.

 Grab  more insights and updates:

Follow me on Instagram: https://instagram.com/the_rhiannonlee

For more FREE resources, templates, guides and information, visit the Designer Resource Hub  on my website ; https://www.rhiannon-lee.com/

Get your interior design studio running on AI in 6 weeks.

admin, client follow-ups, content, and operations handled.
Six live build sessions.

You leave with a working system, not more mess.

Studio BUILD  ( now open) 

Remember to subscribe to the podcast and leave a review. Your feedback helps me continue providing valuable content to aspiring interior designers. Stay tuned for more episodes filled with actionable insights and inspiring conversations. 

Thank you for your support!

Transcript

0:29

Welcome to Designing Success from Study to Studio. I'm your host, Rhiannon Lee, founder of the Oleander and Finch Design Studio. I've lived the transformation from study to studio and then stripped it bare and wrote down the framework so you don't have to overthink it. In this podcast, you can expect real talk with industry friends, community, connection, and actionable tips to help you conquer whatever's holding you back. Now let's get designing your own success we're past talking about if you should or shouldn't use AI, we are absolutely using it. Even if you are staunchly anti-AI, and you'd likely not be listening to this podcast if you were, but it's probably in different tech and software that you haven't even realized. So we're using Chat GPT, we're using Claude, we're using maybe Gemini. You're using Canva. Designers are using transcription tools at in-home consults. We use AI inside of Google and Microsoft and Notion, Adobe, project management software, probably about 50 or 60 other places where we don't even particularly think of it as using AI. You didn't go there for artificial intelligence, but it is embedded inside of that particular software, the question that's been playing on my mind all week is what happens when your client starts asking you questions about how you use AI? I think that's absolutely going to happen if it hasn't already. It's happened to me and I've had to reply and I have been caught feeling really uncomfortable, or if I'm honest And I've had to reply, and if I'm honest, I wasn't ready. So I feel like my reply was very awkward at best, and I don't want that for you. I think, like, if we know these questions are coming, let's just get them out on the table, see if we can get answers. I think that's the best course of action is to really think about your internal AI policies, your external policies, feeling confident so that if somebody comes to you... So that if a client comes to you and actually asks you about it, you're not sort of like, "Oh, sorry, what?" You're a little bit more like, "Yes, we do. We're very mindful of that, and this is how we approach it." Obviously, this episode is not a call to arms. Like, "Everyone call your lawyer, throw your laptops into the sea. We're about to get caught." It's not that kind of drama. It's just a reminder that as we grow with AI literacy, and I've spoken to so many people recently, um, about this sort of stuff and really started to open minds around, "Oh, I didn't think of that," or, "Oh, yeah, maybe they will be interested in that." And I think as our AI literacy grows, so do our clients. Like, they've got their own jobs and their own industries that they are adopting AI into, and they may have internal policies within their job roles that make them think, "Oh, well, hang on, how are you using this stuff within your industry and your job role?" I think clients are becoming increasingly sophisticated about privacy, intellectual property, image rights, confidentiality, and how their information is handled, and they have every right to ask that, obviously. And I think eventually or very soon, in my prediction, I think very soon, people are gonna start to ask you, "Are you using AI in my project? What information are you putting into your AI? Maybe what tools do you use? Are you a Claude user? Do you have an enterprise account?" Or like, "How locked down is your Claude? How locked down is your chat? Are you uploading my floor plans?" Um, you know, I think they will-- I don't think they will be asking you, "Are our meetings being transcribed?" Because if you are nice and transparent, and when you get to the house, you're like, "Look, I'm gonna transcribe this, so I don't drop any details." I always let my clients know they're being recorded as we walk around the house They may even be asking like, "Does your team use a private account or the business account for AI?" I don't know. I'm just... I mean, when I was asked, it was pretty simple. It was just around, are you using... Are you putting my stuff into AI? I think I get the question ahead of time, so I've had this question about six months ago because design clients that are working with me are also exposed to the way that I present in social media and in... Oh, they don't really get my emails. That would be weird. They're excluded from my, that section of the emails that I send out. But you can see from space that I'm very much involved in the AI world, and if I am your designer and I'm in your home, it probably stand to reason that you're going to ask me straight out, "Oh, okay, well, you look like a super user of AI, so how are you actually using our data? And, you know, do you have anything to protect us or you or whatever?" So that question came to me quite a while ago, and it's really been playing on my mind because I think if it's come to me early, it's because of my AI talk and AI influence, and the next step is it's gonna become quite mainstream or, you know, easily expected from your clients out there as well. So they wanna know where does my information go? They might wanna know if you have an internal policy. I mean, they might have an internal policy, as I say, in their industry, so they're thinking, "Okay, well, what are you doing to handle my private info? Have you thought about it? Are my photographs being uploaded into image generators?" Like, they'll start to know more about that stuff. They're getting just as savvy as you are. So you might reflect and think about where you were at this time 12 months ago with your AI literacy, and you're like, "Oh, well I, you know, barely used it to write a caption," or, "I was just getting started and I'm feeling really confident now." I'm starting to speak to a lot more designers who are starting to feel confident, and I love that so much. But so are your clients, right? 'Cause they're also exposed to what's possible So the reason for today's episode is I just wanna put the devil's advocate out there to be like, "Do you think you could answer those questions today? Do you feel confident?" 'Cause I wasn't, I wasn't ready, and I knew what I had sort of... Like, I, I certainly felt confident in answering the how and the why and what I use, and I also knew how the platforms perform and what type of model I had and why, and I had read some of the terms and conditions. So I wasn't... I hadn't prepared an answer. I felt like it, it threw me a little bit, and it... I was on the back foot, and I stumbled and fumbled through it, but I managed to get through it because I had the background education. But I feel like, and I may be wrong here, but I feel like there are a lot of people out there who are like, "Oh, I don't really know what it does with it," or, "I, you know, whilst I do put things in." And I have instructed many people to use the likes of Penelope, that's a proposal generator, and, you know, let's do this, let's do that, and that sits inside of a GPT, so that gives its own different layer of privacy. Anyway To my mind, I don't actually think the answer of, "Oh, don't worry, I have my settings turned off so I can't train your details, don't train the data for everyone," that doesn't quite cut it 'cause that is not enough in terms of privacy. And I feel like... And I'm curious of whether or not you could actually explain what your studio does. And designers who are getting ahead of this, who start to, you know, unpack it in their mind, understand it, get ahead of this conversation, are going to be in a much stronger position than the designers who wait until the client asks you, and then you're on the spot and you're gonna be just as awkward as I was. And that conversation the first time, if you're not prepared, you're completely unprepared, it's gonna throw you. And you may not have an answer, or the answer that you give may not give the client any confidence whatsoever. Like, they might be a bit shirty about it 'cause they're like, "Well, have you even thought about this? This is my home address. This is my, where my bedrooms are located. Like, th- this is sensitive to me, it's important to me." And we are very respectful of that in terms of internally in our documentation. We're not out taping people's floor plans to a front window of our residential retail space that we have, or, you know, we don't leave files open for other clients to sit down and look at their material over the top of somebody else's, like, before photos of their living room, et cetera. Like, we know what to do here A lot of the conversations start with worrying about designers' intellectual property. I feel like I have been having that conversation a lot, and rightfully so. Especially, you know, creativity and artificial intelligence is a hugely uncomfortable conversation because we don't want artificial intelligence using our creative IP. That is our art form. That is, i- in essence, the soul of design, and we do not wanna be outsourcing that or allowing AI to take control of that in any way. I have talked a lot about protecting your IP. I've given thousands and thousands of designers an AI clause, um, and a conversation to start around in contracts, around asking clients not to feed your IP into their private AI. You know, making sure there's an email that says, "Hey, we don't allow that here. If you need changes, you need to come back to me." If you haven't got that contract before or you haven't looked at that clause, you can DM me over on Instagram the_rhiannonlee. DM me the word contract and you'll get a copy. It's not a legal contract, it's just a conversation that says, "Hey, these are the things I talk to my clients about. I don't want them putting my concepts into any AI tool," and I let them know that. So we do need to be having those conversations I don't think a client should ever automatically be able to take your drawings, your concepts, your elevations, your renders, your creative work, your way of curating the world, and feed it into AI to generate variations without considering the rights attached to that work. So that work obviously solely belongs to you and your studio While I was researching the si- while I was researching that side of the conversation, I kept coming back to something that was making me kind of uncomfortable. We're asking clients to respect our intellectual property. That's fantastic. We should be. But what are we properly doing to respect theirs? What are we doing with their information? Like, yeah, those before photos, photographs of their house, their floor plans, their emails. People drop a lot of stuff. I have been a marriage counselor more than once. I know you guys are feeling me on that. Their budget, so financially what they can bring to the table. S- some people, rightfully so, again, are very sensitive about that. The meeting transcripts. Whilst in that meeting, their kids can be at home and recorded. Their... They could be having, you know, sidebar family arguments and things that they just probably don't want in, they may not want in AI. There's a lot of information regarding their family or connected to their family. There can be commercial information, specifically with commercial projects, absolutely. Um, supplier quotes that they've been given. The end-to-end product documentation is so detailed. We know that. You guys are excellent at your jobs. But suddenly that makes the conversation get bigger than, should I just turn the model off in ChatGPT and stop it from training on this data? Well, yeah, you should do that, but that's not the solution to thinking about AI, data protection, cybersecurity, and your responsibility for your clients Throwing in like a, you know, a little line at the bottom of documentation saying, "I use AI," great, be transparent. But that could mean 100 different things. I've just, like, listed a whole bunch of different applications that have internal AI. It's probably the first misconception we need to get rid of. Just y- you cannot get out of this just by saying, "Oh, I use AI." Because someone might tell me, uh, and they do tell me, like, I'll be coaching a private coaching client will say, "I barely use AI with clients. I use it for myself, for my socials, blah, blah, blah, but I barely use it with clients." And then I ask a few more questions like, "Okay, are we recording the transcripts or are we recording consultations?" Tick, yes. "Does something transcribe them?" Yes. So there's obviously it's gone into a system. "Do you use Canva?" "Every day." "Do you use Google, Google Workspace? Do you use Google Workspace? Is it connected to ChatGPT or Claude?" "Yes." "Do you ever use ChatGPT to help word a tricky email?" "Uh, yes." "Okay, so do you upload a PDF and ask Claude to summarize it?" You know, sometimes we're in a hurry, right? Yeah, sometimes. "Does your CRM have AI features?" "Mm, maybe. I haven't used them." Okay, congratulations. We have a fully AI-enabled studio on our hands here, and most of that stuff wasn't because you went into ChatGPT or Claude or Gemini or Copilot or Perplexity, or you had a web browser clipper or a web browser, um, or you had a web browser support assistant run by AI, et cetera. A lot of that stuff was just in tools and software that you've been using for years. They have implemented AI. Things that you're putting into those tools are therefore exposed to AI, and you don't know the data control settings behind what they are doing. You assume, we can all assume, that they're gonna take care, that they're gonna be very careful. But all sorts of things have happened. You know, artifacts that people are making in Claude, once you publish them, they can be What is that word? Artifacts that we're publishing in Claude, they can show up on Google. Like people could get the direct link to your pricing calculator, and even though it won't make any sense to them and it doesn't have client details in it, it will-- it can show up as something... What is that word? What is the word I want where an artifact can show up on Google? So it's not ranking on Google, it's like a word I wanna say, um So your artifact can easily be indexed by Google and just show up randomly somewhere. This is-- I'm never about like freaking you guys out. It's very, very, very unlikely. But if that artifact happened to be a full project for a client, and it had all of their information, and you thought it was safe 'cause you were, you know, doing it up in the artifact section, for whatever reason it was an artifact. I don't know, I'm just making things up. But things can get indexed. It won't be your private conversations, but how would you know? Like, I feel confident that at least seventy percent of people listening to this haven't read the terms and conditions when they signed up to ChatGPT or Claude. So we don't technically know, unless you know. If you know, you know. That feels like so vague, but I-- This is not an anti-AI conversation. I teach people how to use AI in their businesses. I have all these sort of private workshops coming up, and I'm looking at implementation, but safe implementation. I use AI extensively in my business. There are agents, there are all the things. I'm not here to make anyone frightened of uploading a single document. Like, like I said, we're not throwing our lap- we're not throwing our laptops into the sea. We're not afraid of AI. We wanna be educated so that we're ethical with AI, and that is a huge, that is a huge value of me in my business. I'm not just another AI guru, AI expert. I'm always looking at what works for, specifically for interior designers, for our workflows, our processes, and if this is the conversation then we have to have, then guys, this is the conversation we have to have. We have to start thinking, "I'm an interior designer. I wanna be ethical. I wanna be fair to my clients. I want them to be fair to me. This is how I'm gonna move forward." The goal here is to become considerably more deliberate about what we're uploading, where we upload it, why we're uploading it Whether all of that what-- whether all of that matches what we promised our clients, and whether we're educated enough to know what happens when we are putting it in so we can think twice before we do it or be like, "No, I feel pretty good about this because of X, Y, and Z that I know about." Okay. So I'm gonna give you some questions through this episode, and here's one that I want you to think about. Imagine you receive an email from a client, one of those big, long storyboard emails. They've had a big disagreement with their partner about the renovation. There's, you know, some information and questions about the budget. They mention their child's accessibility needs. There's their full name, email address, property details are sitting inside of that email, and you wanna respond carefully, obviously, and give it the considered response that it requires. So you copy the entire email into an AI tool and say, "Help me write a warm response to this." That's a completely normal workflow, right? Probably happens thousands of times every single day across every single industry. But did the AI actually need all of that information? Absolutely not. You could have hit the microphone button and given the AI the gist, like, "I've just received a client email. They're upset because husband doesn't want X. Dependent child needs Y. Help me calmly handle this or reply with empathy." There are no details in that prompt, and you would get a similar thing. Or, "My residential client is feeling overwhelmed about budget changes and accessibility requirements. Help me draft a warm response acknowledging their concerns and proposing a meeting." That's the same, like, really great assistance from your AI tools. You could be copying chunks of, like, paragraphs that you can see have nothing. So it's almost like redacting what is in there and being like, "Here's, you know, bits and pieces of her tone to me, what she was concerned about," but you're automatically taking out those personal details. So there's considerably less information going into the AI, and that's the kind of judgment that I'm really interested in coaching and teaching. I think just knowing what was actually necessary here. Could this be easily redacted or presented as a third-party scenario? Like, "I'm working with a residential client. I've got these ideas. They've got these ideas. They need this from me. I need this from them," and telling the story of that rather than just, you know... And I'm not gonna say lazy, it's just kind of how we learn to use AI, was to upload everything as full context, but that includes personal and private information. Okay? So it took the same amount of energy for me really. Okay, it was a little bit faster to just upload the whole email, but it took a similar amount of energy to do it that way with the sort of redacted, and it felt so much more ethical 'cause I could have an, a conversation with my email GPT or my email agent, um, and yes, you can give these things access directly to your email inbox, so they would have already seen that and helped you reply. It's all about Knowing your governance, your guardrails, your expectations, and making some ethical decisions for yourself. So if I have an agent who reads my inbox, they've probably got a lot stronger guardrails and governance around how I created them, and if I'm up to making agents like that, I'm probably not on a free pl- I'm definitely not on a free plan. I'm probably not on the pro plan. I've probably thought about, mm, for my business going forward, I would actually like a stronger sense of security. I wanna be on like an, a business, an enterprise, a team, and actually know that, um, there are differences. So I'm not gonna read terms and conditions to you now, but I know when I say that I'm working with agents, I am not on the free or p- pro plan of these AI tools. Part of what I do within my business makes it heavily important for me to be on these larger team... on these larger models, on these more expensive and more secure models So I'm gonna talk to you in a little bit about a webinar that I'm running shortly to go through all of this stuff. Like, I doubt that any of you are like, "This is so interesting," because it feels flat and it feels boring, but it is so important. And I'm gonna be going through a lot of this stuff as well as answering questions. I just think it's such a curi- like it's a, a pivotable time in our... It's a pivotal time in our humanity, in our world, in our industries, and in business, and the more you can sort of feel confident about this stuff, the more you can understand it, the easier it is for you to make those decisions. Here's another question I wanna put to you. What actually happens when you delete a chat? Because I think most people intuitively imagine delete chat, information gone. But AI systems can involve several different layers of processing and storage. There's the conversation that we see, like the thread or the conversation that you're talking in. There may be files attached that are retained. There may be logs at server level. There could be retention periods, like they might hold onto things. Like I know there is a help with Chat and Claude, and sometimes if you go to them and ask, they're like, "Oh, we hold the data for 90 days or whatever, so we can pull it back out if you've lost something." And I'm like, "Oh, hang on. What, what's this?" So, you know, do you know that stuff? I don't know if you do. There could be workplace permissions. There could be integrations or connected services that, you know, when you're connecting Google and Claude and all these things, like is it maybe retained on one and not on the other? Critically, they could be processing your information Critically, processing your information isn't automatically the same thing as training the underlying AI model. The two ideas get mashed together constantly, so when someone says, "Don't upload anything because AI trains on everything," that's too simplistic. But equally, "I've turned training off, therefore I can upload whatever I like," is also too simplistic. That's-- It's nuanced. It's not like that. The question could be what actually happens to this information in the particular product or the particular account using this particular feature? Do you know that, the answer to that? Your account type, your account type really matters. It's something that I want designers to understand. We can't just say Chat GPT is safe or Claude is private or Gemini doesn't train on the information. The answer will depend entirely on what product and account you're using as a consumer. So are you just a direct consumer? Are you taking the free model? Are you on a business plan, an enterprise plan? Are you using APIs, workspaces, teams? Different products come with different promises around training, retention, admin, access, data handling. The same platform can have totally different environments and expectations based on the model that you've bought or the... What is it called? Based, based on the pricing tier that you've agreed to, which means Which means one of the questions every studio should eventually be able to answer if someone asks you is, which AI tools have we actually approved for client work? So underneath that, you want to say, "Okay, well, what client work have I approved for this? Is it a tool that I'm happy to use to brainstorm Instagram hooks?" That's not the same tool that you would automatically place unredacted client contracts. Yeah. That... But maybe it is, depending on your model. This is what you have to think about. Then there's your team. If you've got staff, this gets really interesting really quickly. You might be incredibly careful. You might do some training with me. You might have had one-on-ones. You might have locked it all down. You might have been really quite savvy in this area or done some external research. And then what about your junior designer? What about your offshore VA? Your copywriter? Like, maybe you've got someone just doing the website and you're working together, but are they pop- pumping intel about you into AI that's unprotected? If you've got a third-party rendering contractor or a bookkeeper, you don't know... Like, you can have bookkeepers who don't know as much about AI as they think they do, but they can see that you can connect MYOB, QuickBooks, Xero, et cetera. Are they connecting Xero through their AI with your private data? Do they have an internal policy? Have you signed something? Like, these are big, big questions, right? What accounts are they using? Are they copying client information into their personal account, a business, an enterprise, a freebie? S- so many people are still on the free versions of these tools, yeah? So it's definitely time we started talking. You'll see a lot of people's note takers coming into meetings these days. You know, they're automatically transcribing. I don't see a lot of consent forms changing hands. I think it's really wild, actually, that I am very, very transparent and clear if you're working with me. This is being recorded. These are the reasons. This is the benefits. This is what I do. I use this particular type of AI at this level so that your stuff is more protected, et cetera. Um, you know, as much as I pay for all that extra stuff, I also use it when I'm private coaching. If I'm designing you an app, I'm going to use Fable 5, the king of Claude models, so that you can use some of my tokens to get a way better thing than if you're using your freebie, for example. So, um, it matters to me to have this higher level of security, but also higher level of use Browser extensions are such a danger zone here. Like, are they reading information? Are you logging into things? What happens when the project finishes? Like, at the end of your project, does anybody delete what was there? Like, okay, so that delete question that I asked before, imagine that the terms and conditions say it's retained for 90 days, but if you're at the end of the project, you've got everything you need in your CRM and final documentation, it is best practice to delete that conversation so within 90 days it then no longer hovers around the internet in the ethos, in the magic cloud. I, just side note, am so impressed with myself building... You know, being so ahead of this stuff and building all these GPTs and doing all of this stuff when I remember it took me, it's embarrass- an embarrassing long time to actually understand what the cloud was. Like, just being like, "But where's the cloud?" Like, "Yeah, okay, but where does it live? Like, what does it look like?" And I guess that's 'cause it was so long ago, but now I, it, it makes me giggle thinking of all of these things that I'm doing to be, and being ahead, especially at my age, and then how foreign the concept of just a cloud was to me. Anyway, let's get back. I'm making myself look like an idiot. Okay. So- What I'm describing, like not actually knowing what our third-party contractors and stuff are doing, is what some people call shadow AI, okay? So if you hear that term, people sometimes call it shadow AI, and it means that AI tools are being used within the business without the business necessarily knowing, approving, controlling that use. So it's like third-party use shadowing. And again, this isn't necessarily because somebody has done something terrible. If you are my contractor, hi Carly, if you've used AI, I'm not coming for you. I know you didn't do anything on purpose. It's often because nobody just created a policy or a rule. Nobody started the conversation. And I want to be so fair about this. Why would they have started the conversation? This is brand new to our brains. Like, these, these are all things that didn't exist even s- even six to 12 months ago when I was building Penelope. This wasn't a conversation that I was really diving into as much as I have done in the last sort of, yeah, six or nine months. So since building Penelope and thinking, "Okay, well, I'm selling this technology that auto-creates... I'm selling a technology that is an end-to-end system, will write proposals in like 10 minutes instead of a whole day," and it's beautiful. If you're interested, you can head to my website and check it out. But it's... And it's fabulous, and it still works. If you're interested, you can hit me up anytime. I'll pop a link in the show notes. But it is I'll pop a long-- I'll pop a link in the show notes. But it is a risk still, right? Because I'm asking you to put a full transcription of an in-home consult into ChatGPT specifically because Penelope is a GPT. Now, GPTs are locked down in a different way, and I'm not gonna get into all the backend stuff. So I'm-- I've done my research. I'm feeling good. But it brings me back to the AI clause that we talked about, because these are all things that are quite new conversations. If you downloaded my contract clause to start thinking about setting expectations for your clients around them feeding your AI work in, that clause is just the tip of the iceberg of the conversation that I actually wanna have. Because in the clause, it essentially helps you to communicate, "Here's what we promise." Your operations determine whether you can actually keep that promise, and that's where it becomes a business system conversation. I need a little button in this studio where I can hit, like, a little nerd alert button every time I talk about processes and systems. But they really, really matter, and they really matter to me because I see on the other side how good documenta- how good documented businesses connect with AI. Like, the girls coming out of Studio Build, it's a whole new world for them, and I love it. I also see, you know, just focusing on one thing. You're not gonna have to talk about contracts and clauses and internal policies for AI forever. You have to do it once, make some strong decisions, and sorry, like, big girl pants on. You are the founder of this business. You have to make the decision, write it down, draw a line in the sand, and then just review it every... I'd say probably every quarter because things are moving fast, so you just wanna make sure that you've got everything under control And we can't just say, uh, you know, you can't just say to your clients, "Oh yeah, I've thought about this, I've got it under control." Because if your contract says personally identifiable client information won't be uploaded into AI, who's gonna make sure that that happens? Is it just you? Like, maybe it is, and that's fine. But what counts as personally identifiable? Have you been careful to explain that? And does your contractor know? Have you gone and told your Carly? Does your VA know? You know, have you gone and told your Jess? What about meeting transcripts? Are they, like, just separate because you already get-- You know, do you get written consent when you go in? Do you get verbal consent and that consent is transcribed so you have it? Um, are you even asking that? I have always just assumed that people know when they're going into a house that you'd say, "Hey, I'm actually gonna pop this recording technology on now because I wanna capture every detail, and this is the reason why. Are you comfortable with that?" And actually getting people's consent. Consent matters. As a mum of three boys, I am s- seriously speaking about consent every single day, and it really, really is important. If somebody doesn't want you to do that, you gotta go old school. You've gotta get your notepad out, and you've gotta take really good notes in that in-home consult because they have every right to say no to having it transcribed. They have also got a right to say, "Yes, you can transcribe it, but no, I don't want you to put it into AI," and these are the, the things that we are gonna be talking about really soon What about putting in like a floor plan with the title block? Or what happens when someone, uh, like a- another team member, even yourself, you're sometimes guilty of this, is rushing at 2:45 to get to school pick-up on a Friday, they need AI to summarize something really quickly so that we could actually reply to the email before we go. Are we just gonna break our own policies, or like do we have an approval tool, a redaction process? Have you got a restricted data list, like things that absolutely cannot go into AI? Do we have a human review checkpoint? Like is there a process for who looks over this crap sometimes, slop I should say. Who looks o- over this AI slop coming out and supposedly going to be our caption for Tuesday? Like is there a human review checkpoint, because that is so important. Or is the entire AI policy currently stored inside of your head? Psych, there was never any policy. Don't come and pretend to me that you have a walking AI policy s- currently stored inside of your head. This is one of those examples where like the procedure doesn't live in the founder's head, it just doesn't exist, and that's so fine. There's no shade. This stuff is really new, and having a note in a CRM that just says, "Oh, Rhiannon knows what not to upload," that's not a business system. It's not sustainable and it's not, you know... I'm, I'm not a lawyer. I'm never gonna get into that conversation of what will and won't stand up in court. The legislation has not caught up with AI. The law hasn't caught up with AI. God only knows what courtrooms, you know, what courtroom conversations are going to be had in the future and what will and won't hold up. I'm not about that. I'm not talking to you from that lens. I'm talking to you as a founder of a, as a founder of a successful business, as somebody who works with design clients, who knows your workflows, your tech, what you're doing, what will help you, and how to get you to be able to access and unlock all of those AI wins without putting our head in the sand about this side of it as well. Because ethical AI is where I stand, and I fiercely believe in keeping creativity out of AI, not putting our, not putting anything that makes us amazing designers into these tools. They are great for so many things, but they are not great for that This is where I think the opportunity is. I don't want designers waiting for a client to send them an email saying, "Hey, can you please provide your policy on use- on the use of generative AI and client information?" And then suddenly you're Googling frantically or worse than that, you've just gone to Claude and said, "Can you write me a policy for the internal use of generative AI and client information?" Um, I would much rather in that scenario, you were able to go back and just say, "Look, absolutely. Here's how we use AI. Here's what we don't put into our AI tools. Here's how we protect identifiable information. Here's where the human review level sits. Um, here's how our team and contractors work within those roles. Here's our approach to third-party software that we use, and we, you know, we have someone that runs through that or we..." You know, you can take the terms and conditions of Canva, put it into Claude, ask for the key takeaways, ask it, ask it to look for any, um, anything that you should be alarmed about as a user who will be putting client stuff in. Like, you can do this work. It's not sexy and it's not fun, but if you're curious, if you're thinking, "I don't know the answer to any of the questions this chick is asking me today," then you can do a little bit of digging in that way. I shared in my series on Instagram this week, I'm doing one tip and one truth on AI every day. You can check them out in the highlights, um, if you wanna go back and, like, learn a, a couple of those tips and tricks and truths. A couple of those tips and truths. Uh, and one of them that I had shared is just how to, like, find all four keywords that will pretty much help you unpack the terms and conditions of your large language models. So that might be helpful. You can go back and check that out on Instagram. If you can reply like that, like this is actually our approach, that feels so different. It's not awkward, it's not defensive, it's totally professional. You're not gonna bother me. You're not like, "Sit down, let's talk about the data policy." But increasingly, I think this is gonna become a part of what a well-run modern studio looks like. So here are the questions I told you this episode is full of questions, and the answers will be addressed in my free webinar on the 24th of... The answers are something that I will be going into a lot more in my webinar that I'm running on the 24th August, but here are the questions I want you to leave... Here are the questions I wanna leave with. Don't try to answer them all right now. You probably won't know them. I just wanna... I just want you to listen, if you're walking the dog, if you're out there, going to school pick-up, you're driving, you're mopping the floors, however you listen to Designing Success, I want you to listen to these questions and just take note of your physicality. Like which ones make you squirm? This'll be fun. Okay. What AI-enabled tools currently touch client information in my studio? Which ones have I deliberately approved and which ones just snuck into my workflow? Like I've been using them for years, and now they've added additional AI, and I didn't even know. What client information am I currently uploading? How much of that information does AI actually need? What is necessary? Do I know whether my account is consumer, business, or enterprise? Do I understand what happens to uploaded files on that account? Do I know how long information is retained with my particular pricing tier? What happens when I delete something in my AI platform? Who else can use my workspace? Like who has access to my AI? What AI tools are my staff and my contractors using? What... Are client meetings being recorded and transcribed? Am I uploading intellectual property I don't actually own? That could be your photographer's images, like anything like that. Just think about that. Am I uploading intellectual property I myself don't actually own? What have I promised clients in my contract and privacy documentation? The privacy policy on my website, what does it say? Does it include AI? Probably not Do my actual workflows match those promises? And most importantly, if a client asks me about this today or tomorrow, this afternoon, next in home, can I confidently answer them This is exactly what I'm talking about in the webinar I'm running. It's a free session called What's Really Happening When You Upload Files to AI: The Truth About Client Data, Intellectual Property, and AI for Interior Designers. We're gonna get properly into this. We're gonna unpack it. I promise it won't just be questions. It will also be just a level of education around in the research that I've done, this is what I've found, this is what I think you need to know. It's not going to be a, like, scary legal chat about everything that you've done wrong. And when we know better, we do better. We're not going back into the past and just going to, like, stay awake all night every night being like, "I think I've un-uploaded this thing." What's done is done. You could go find that chat and delete it if you would like to. Um, you know, and then it may take 60, 90, 120 days, whatever your contract says with your platform. But you can only go forwards better, if that makes sense. Like, once we know this stuff, we can write an internal policy document. You can start to let people around you know what your expectations are. We're getting properly into it. We're going to understand the difference between processing, training, why your account type matters, what information you should think twice about sharing. Um, you know, what you should and shouldn't be sharing with AI. You know, floor plans, yes or no? Emails, yes or no? Marketing transcripts, yes or no? Client photographs, good, bad, yes, no? Supplier information, is that okay? Intellectual property that you own versus that you don't own. What should you be checking in the tools that you're already using? Like, where can you find the answers? And how to start creating some sensible boundaries inside your studio without becoming terrified of AI or producing 150-page policy that no one's ever gonna read. You didn't read the one from Claude or ChatGPT, so why are they gonna read the one from, um, designco.company or whatever your design brand is? I am absolutely not interested in terrifying everybody. Like, if you come along to the webinar, it's not a lecture. It's not going to... It's just education that I think we-- it's conversation that we all need to have. It's education piece that I think is helpful because confidence is everything and edu- like, just taking 45 minutes out of your day to be at a webinar or watch the replay. If you sign up to this webinar and you can't attend live, I will send you the replay. Just being able to do that can help you use AI in a way that feels ethical, feels confident, and feels like you've ticked your responsibility to learn a little bit more about the boring background s- um, stuff that you've signed up for. You may come out of that and think, "And I don't work for these guys." I say it all the time, but I don't work for Silicon Valley. I get no kickbacks. If you go from a pro plan up to a business or enterprise and you lock things down a lot tighter, that is a business decision that you have made. I don't work for them. I'm not getting a kickback. Um, I just want the businesses underneath AI, that are using AI, to be able to catch up, to feel confident, and most importantly, when you are on site, I want you to start Like start this conversation. Like if it comes from you, they're gonna ask you less questions. If you're confident with your clients before they're the ones starting the questions with you, and just have something that's like, "Yep, we use AI. Here's how we handle it responsibly." That is a way better position than being caught completely off guard when somebody says-- when somebody asks you these questions. Okay? So I'm gonna put the registration link in the show notes. Please come along or watch the replay. Bring all the AI questions you've currently... Bring all the AI questions that you're freaking out about, wondering whether you could... You know, who you're gonna ask. Come and ask me. I may not know the answers. This stuff moves so fast, faster than all of us, right? I'm not the expert, but I'm fully emerged in this world all the time through the lens of interior design and AI, so if anyone's going to know, it's likely gonna be me, and if it's not me, I'm likely gonna wanna find the answer for us both. I suspect we're gonna be talking about this stuff a lot more in the future, so we might as well just tick it off, get it done. I'm really excited. It's Monday the 24th of August. It's at 12:30 PM Australian Eastern Standard Time. The registration is in the show notes. You can also just contact me on Instagram, and I will send you a link. If you're on my email list, you will have seen it. There are hundreds of people already registered, and it's gonna be a really good conversation full of lots of different types of in- design firms and different ways of using AI. So I cannot wait Okay. Well, sorry if I freaked you out. If you have gone for your walk with the dog today and you're thinking, "Oh my God, I haven't thought of any of this stuff, and I'm just like a full AI cowboy," I'll see you in the webinar. Okay, bye for now. gonna start asking how you use AI, and I have a feeling we're not quite ready to answer. So we're past talking about if you should or shouldn't use AI, we are absolutely using it. Even if you are staunchly anti-AI, and you'd likely not be listening to this podcast if you were, but it's probably in things and in different tech and software that you haven't even realized. So we're using Chat GPT, we're using Claude, we're using maybe Gemini. You're using Canva. Designers are using transcription tools at in-home consults. We use AI inside of Google and Microsoft and Notion, Adobe, project management software, probably about 50 or 60 other places where we don't even particularly think of it as using AI. You didn't go there for artificial intelligence, but it is embedded inside of that particular software, inside of that particular software. The question that's been playing on my mind all week is what happens when your client starts asking you questions about how you use AI? I think that's absolutely going to happen if it hasn't already. It's happened to me and I've had to reply and I have been caught feeling really uncomfortable, or if I'm honest And I've had to reply, and if I'm honest, I wasn't ready. So I feel like my reply was very awkward at best, and I don't want that for you. I think, like, if we know these questions are coming, let's just get them out on the table, see if we can get answers. I think that's the best course of action is to really think about your internal AI policies, your external policies, feeling confident so that if somebody comes to you... So that if a client comes to you and actually asks you about it, you're not sort of like, "Oh, sorry, what?" You're a little bit more like, "Yes, we do. We're very mindful of that, and this is how we approach it." Obviously, this episode is not a call to arms. Like, "Everyone call your lawyer, throw your laptops into the sea. We're about to get caught." It's not that kind of drama. It's just a reminder that as we grow with AI literacy, and I've spoken to so many people recently, um, about this sort of stuff and really started to open minds around, "Oh, I didn't think of that," or, "Oh, yeah, maybe they will be interested in that." And I think as our AI literacy grows, so do our clients. Like, they've got their own jobs and their own industries that they are adopting AI into, and they may have internal policies within their job roles that make them think, "Oh, well, hang on, how are you using this stuff within your industry and your job role?" I think clients are becoming increasingly sophisticated about privacy, intellectual property, image rights, confidentiality, and how their information is handled, and they have every right to ask that, obviously. And I think eventually or very soon, in my prediction, I think very soon, people are gonna start to ask you, "Are you using AI in my project? What information are you putting into your AI? Maybe what tools do you use? Are you a Claude user? Do you have an enterprise account?" Or like, "How locked down is your Claude? How locked down is your chat? Are you uploading my floor plans?" Um, you know, I think they will-- I don't think they will be asking you, "Are our meetings being transcribed?" Because if you are nice and transparent, and when you get to the house, you're like, "Look, I'm gonna transcribe this, so I don't drop any details." I always let my clients know they're being recorded as we walk around the house They may even be asking like, "Does your team use a private account or the business account for AI?" I don't know. I'm just... I mean, when I was asked, it was pretty simple. It was just around, are you using... Are you putting my stuff into AI? I think I get the question ahead of time, so I've had this question about six months ago because design clients that are working with me are also exposed to the way that I present in social media and in... Oh, they don't really get my emails. That would be weird. They're excluded from my, that section of the emails that I send out. But you can see from space that I'm very much involved in the AI world, and if I am your designer and I'm in your home, it probably stand to reason that you're going to ask me straight out, "Oh, okay, well, you look like a super user of AI, so how are you actually using our data? And, you know, do you have anything to protect us or you or whatever?" So that question came to me quite a while ago, and it's really been playing on my mind because I think if it's come to me early, it's because of my AI talk and AI influence, and the next step is it's gonna become quite mainstream or, you know, easily expected from your clients out there as well. So they wanna know where does my information go? They might wanna know if you have an internal policy. I mean, they might have an internal policy, as I say, in their industry, so they're thinking, "Okay, well, what are you doing to handle my private info? Have you thought about it? Are my photographs being uploaded into image generators?" Like, they'll start to know more about that stuff. They're getting just as savvy as you are. So you might reflect and think about where you were at this time 12 months ago with your AI literacy, and you're like, "Oh, well I, you know, barely used it to write a caption," or, "I was just getting started and I'm feeling really confident now." I'm starting to speak to a lot more designers who are starting to feel confident, and I love that so much. But so are your clients, right? 'Cause they're also exposed to what's possible So the reason for today's episode is I just wanna put the devil's advocate out there to be like, "Do you think you could answer those questions today? Do you feel confident?" 'Cause I wasn't, I wasn't ready, and I knew what I had sort of... Like, I, I certainly felt confident in answering the how and the why and what I use, and I also knew how the platforms perform and what type of model I had and why, and I had read some of the terms and conditions. So I wasn't... I hadn't prepared an answer. I felt like it, it threw me a little bit, and it... I was on the back foot, and I stumbled and fumbled through it, but I managed to get through it because I had the background education. But I feel like, and I may be wrong here, but I feel like there are a lot of people out there who are like, "Oh, I don't really know what it does with it," or, "I, you know, whilst I do put things in." And I have instructed many people to use the likes of Penelope, that's a proposal generator, and, you know, let's do this, let's do that, and that sits inside of a GPT, so that gives its own different layer of privacy. Anyway To my mind, I don't actually think the answer of, "Oh, don't worry, I have my settings turned off so I can't train your details, don't train the data for everyone," that doesn't quite cut it 'cause that is not enough in terms of privacy. And I feel like... And I'm curious of whether or not you could actually explain what your studio does. And designers who are getting ahead of this, who start to, you know, unpack it in their mind, understand it, get ahead of this conversation, are going to be in a much stronger position than the designers who wait until the client asks you, and then you're on the spot and you're gonna be just as awkward as I was. And that conversation the first time, if you're not prepared, you're completely unprepared, it's gonna throw you. And you may not have an answer, or the answer that you give may not give the client any confidence whatsoever. Like, they might be a bit shirty about it 'cause they're like, "Well, have you even thought about this? This is my home address. This is my, where my bedrooms are located. Like, th- this is sensitive to me, it's important to me." And we are very respectful of that in terms of internally in our documentation. We're not out taping people's floor plans to a front window of our residential retail space that we have, or, you know, we don't leave files open for other clients to sit down and look at their material over the top of somebody else's, like, before photos of their living room, et cetera. Like, we know what to do here A lot of the conversations start with worrying about designers' intellectual property. I feel like I have been having that conversation a lot, and rightfully so. Especially, you know, creativity and artificial intelligence is a hugely uncomfortable conversation because we don't want artificial intelligence using our creative IP. That is our art form. That is, i- in essence, the soul of design, and we do not wanna be outsourcing that or allowing AI to take control of that in any way. I have talked a lot about protecting your IP. I've given thousands and thousands of designers an AI clause, um, and a conversation to start around in contracts, around asking clients not to feed your IP into their private AI. You know, making sure there's an email that says, "Hey, we don't allow that here. If you need changes, you need to come back to me." If you haven't got that contract before or you haven't looked at that clause, you can DM me over on Instagram the_rhiannonlee. DM me the word contract and you'll get a copy. It's not a legal contract, it's just a conversation that says, "Hey, these are the things I talk to my clients about. I don't want them putting my concepts into any AI tool," and I let them know that. So we do need to be having those conversations I don't think a client should ever automatically be able to take your drawings, your concepts, your elevations, your renders, your creative work, your way of curating the world, and feed it into AI to generate variations without considering the rights attached to that work. So that work obviously solely belongs to you and your studio While I was researching the si- while I was researching that side of the conversation, I kept coming back to something that was making me kind of uncomfortable. We're asking clients to respect our intellectual property. That's fantastic. We should be. But what are we properly doing to respect theirs? What are we doing with their information? Like, yeah, those before photos, photographs of their house, their floor plans, their emails. People drop a lot of stuff. I have been a marriage counselor more than once. I know you guys are feeling me on that. Their budget, so financially what they can bring to the table. S- some people, rightfully so, again, are very sensitive about that. The meeting transcripts. Whilst in that meeting, their kids can be at home and recorded. Their... They could be having, you know, sidebar family arguments and things that they just probably don't want in, they may not want in AI. There's a lot of information regarding their family or connected to their family. There can be commercial information, specifically with commercial projects, absolutely. Um, supplier quotes that they've been given. The end-to-end product documentation is so detailed. We know that. You guys are excellent at your jobs. But suddenly that makes the conversation get bigger than, should I just turn the model off in ChatGPT and stop it from training on this data? Well, yeah, you should do that, but that's not the solution to thinking about AI, data protection, cybersecurity, and your responsibility for your clients Throwing in like a, you know, a little line at the bottom of documentation saying, "I use AI," great, be transparent. But that could mean 100 different things. I've just, like, listed a whole bunch of different applications that have internal AI. It's probably the first misconception we need to get rid of. Just y- you cannot get out of this just by saying, "Oh, I use AI." Because someone might tell me, uh, and they do tell me, like, I'll be coaching a private coaching client will say, "I barely use AI with clients. I use it for myself, for my socials, blah, blah, blah, but I barely use it with clients." And then I ask a few more questions like, "Okay, are we recording the transcripts or are we recording consultations?" Tick, yes. "Does something transcribe them?" Yes. So there's obviously it's gone into a system. "Do you use Canva?" "Every day." "Do you use Google, Google Workspace? Do you use Google Workspace? Is it connected to ChatGPT or Claude?" "Yes." "Do you ever use ChatGPT to help word a tricky email?" "Uh, yes." "Okay, so do you upload a PDF and ask Claude to summarize it?" You know, sometimes we're in a hurry, right? Yeah, sometimes. "Does your CRM have AI features?" "Mm, maybe. I haven't used them." Okay, congratulations. We have a fully AI-enabled studio on our hands here, and most of that stuff wasn't because you went into ChatGPT or Claude or Gemini or Copilot or Perplexity, or you had a web browser clipper or a web browser, um, or you had a web browser support assistant run by AI, et cetera. A lot of that stuff was just in tools and software that you've been using for years. They have implemented AI. Things that you're putting into those tools are therefore exposed to AI, and you don't know the data control settings behind what they are doing. You assume, we can all assume, that they're gonna take care, that they're gonna be very careful. But all sorts of things have happened. You know, artifacts that people are making in Claude, once you publish them, they can be What is that word? Artifacts that we're publishing in Claude, they can show up on Google. Like people could get the direct link to your pricing calculator, and even though it won't make any sense to them and it doesn't have client details in it, it will-- it can show up as something... What is that word? What is the word I want where an artifact can show up on Google? So it's not ranking on Google, it's like a word I wanna say, um So your artifact can easily be indexed by Google and just show up randomly somewhere. This is-- I'm never about like freaking you guys out. It's very, very, very unlikely. But if that artifact happened to be a full project for a client, and it had all of their information, and you thought it was safe 'cause you were, you know, doing it up in the artifact section, for whatever reason it was an artifact. I don't know, I'm just making things up. But things can get indexed. It won't be your private conversations, but how would you know? Like, I feel confident that at least seventy percent of people listening to this haven't read the terms and conditions when they signed up to ChatGPT or Claude. So we don't technically know, unless you know. If you know, you know. That feels like so vague, but I-- This is not an anti-AI conversation. I teach people how to use AI in their businesses. I have all these sort of private workshops coming up, and I'm looking at implementation, but safe implementation. I use AI extensively in my business. There are agents, there are all the things. I'm not here to make anyone frightened of uploading a single document. Like, like I said, we're not throwing our lap- we're not throwing our laptops into the sea. We're not afraid of AI. We wanna be educated so that we're ethical with AI, and that is a huge, that is a huge value of me in my business. I'm not just another AI guru, AI expert. I'm always looking at what works for, specifically for interior designers, for our workflows, our processes, and if this is the conversation then we have to have, then guys, this is the conversation we have to have. We have to start thinking, "I'm an interior designer. I wanna be ethical. I wanna be fair to my clients. I want them to be fair to me. This is how I'm gonna move forward." The goal here is to become considerably more deliberate about what we're uploading, where we upload it, why we're uploading it Whether all of that what-- whether all of that matches what we promised our clients, and whether we're educated enough to know what happens when we are putting it in so we can think twice before we do it or be like, "No, I feel pretty good about this because of X, Y, and Z that I know about." Okay. So I'm gonna give you some questions through this episode, and here's one that I want you to think about. Imagine you receive an email from a client, one of those big, long storyboard emails. They've had a big disagreement with their partner about the renovation. There's, you know, some information and questions about the budget. They mention their child's accessibility needs. There's their full name, email address, property details are sitting inside of that email, and you wanna respond carefully, obviously, and give it the considered response that it requires. So you copy the entire email into an AI tool and say, "Help me write a warm response to this." That's a completely normal workflow, right? Probably happens thousands of times every single day across every single industry. But did the AI actually need all of that information? Absolutely not. You could have hit the microphone button and given the AI the gist, like, "I've just received a client email. They're upset because husband doesn't want X. Dependent child needs Y. Help me calmly handle this or reply with empathy." There are no details in that prompt, and you would get a similar thing. Or, "My residential client is feeling overwhelmed about budget changes and accessibility requirements. Help me draft a warm response acknowledging their concerns and proposing a meeting." That's the same, like, really great assistance from your AI tools. You could be copying chunks of, like, paragraphs that you can see have nothing. So it's almost like redacting what is in there and being like, "Here's, you know, bits and pieces of her tone to me, what she was concerned about," but you're automatically taking out those personal details. So there's considerably less information going into the AI, and that's the kind of judgment that I'm really interested in coaching and teaching. I think just knowing what was actually necessary here. Could this be easily redacted or presented as a third-party scenario? Like, "I'm working with a residential client. I've got these ideas. They've got these ideas. They need this from me. I need this from them," and telling the story of that rather than just, you know... And I'm not gonna say lazy, it's just kind of how we learn to use AI, was to upload everything as full context, but that includes personal and private information. Okay? So it took the same amount of energy for me really. Okay, it was a little bit faster to just upload the whole email, but it took a similar amount of energy to do it that way with the sort of redacted, and it felt so much more ethical 'cause I could have an, a conversation with my email GPT or my email agent, um, and yes, you can give these things access directly to your email inbox, so they would have already seen that and helped you reply. It's all about Knowing your governance, your guardrails, your expectations, and making some ethical decisions for yourself. So if I have an agent who reads my inbox, they've probably got a lot stronger guardrails and governance around how I created them, and if I'm up to making agents like that, I'm probably not on a free pl- I'm definitely not on a free plan. I'm probably not on the pro plan. I've probably thought about, mm, for my business going forward, I would actually like a stronger sense of security. I wanna be on like an, a business, an enterprise, a team, and actually know that, um, there are differences. So I'm not gonna read terms and conditions to you now, but I know when I say that I'm working with agents, I am not on the free or p- pro plan of these AI tools. Part of what I do within my business makes it heavily important for me to be on these larger team... on these larger models, on these more expensive and more secure models So I'm gonna talk to you in a little bit about a webinar that I'm running shortly to go through all of this stuff. Like, I doubt that any of you are like, "This is so interesting," because it feels flat and it feels boring, but it is so important. And I'm gonna be going through a lot of this stuff as well as answering questions. I just think it's such a curi- like it's a, a pivotable time in our... It's a pivotal time in our humanity, in our world, in our industries, and in business, and the more you can sort of feel confident about this stuff, the more you can understand it, the easier it is for you to make those decisions. Here's another question I wanna put to you. What actually happens when you delete a chat? Because I think most people intuitively imagine delete chat, information gone. But AI systems can involve several different layers of processing and storage. There's the conversation that we see, like the thread or the conversation that you're talking in. There may be files attached that are retained. There may be logs at server level. There could be retention periods, like they might hold onto things. Like I know there is a help with Chat and Claude, and sometimes if you go to them and ask, they're like, "Oh, we hold the data for 90 days or whatever, so we can pull it back out if you've lost something." And I'm like, "Oh, hang on. What, what's this?" So, you know, do you know that stuff? I don't know if you do. There could be workplace permissions. There could be integrations or connected services that, you know, when you're connecting Google and Claude and all these things, like is it maybe retained on one and not on the other? Critically, they could be processing your information Critically, processing your information isn't automatically the same thing as training the underlying AI model. The two ideas get mashed together constantly, so when someone says, "Don't upload anything because AI trains on everything," that's too simplistic. But equally, "I've turned training off, therefore I can upload whatever I like," is also too simplistic. That's-- It's nuanced. It's not like that. The question could be what actually happens to this information in the particular product or the particular account using this particular feature? Do you know that, the answer to that? Your account type, your account type really matters. It's something that I want designers to understand. We can't just say Chat GPT is safe or Claude is private or Gemini doesn't train on the information. The answer will depend entirely on what product and account you're using as a consumer. So are you just a direct consumer? Are you taking the free model? Are you on a business plan, an enterprise plan? Are you using APIs, workspaces, teams? Different products come with different promises around training, retention, admin, access, data handling. The same platform can have totally different environments and expectations based on the model that you've bought or the... What is it called? Based, based on the pricing tier that you've agreed to, which means Which means one of the questions every studio should eventually be able to answer if someone asks you is, which AI tools have we actually approved for client work? So underneath that, you want to say, "Okay, well, what client work have I approved for this? Is it a tool that I'm happy to use to brainstorm Instagram hooks?" That's not the same tool that you would automatically place unredacted client contracts. Yeah. That... But maybe it is, depending on your model. This is what you have to think about. Then there's your team. If you've got staff, this gets really interesting really quickly. You might be incredibly careful. You might do some training with me. You might have had one-on-ones. You might have locked it all down. You might have been really quite savvy in this area or done some external research. And then what about your junior designer? What about your offshore VA? Your copywriter? Like, maybe you've got someone just doing the website and you're working together, but are they pop- pumping intel about you into AI that's unprotected? If you've got a third-party rendering contractor or a bookkeeper, you don't know... Like, you can have bookkeepers who don't know as much about AI as they think they do, but they can see that you can connect MYOB, QuickBooks, Xero, et cetera. Are they connecting Xero through their AI with your private data? Do they have an internal policy? Have you signed something? Like, these are big, big questions, right? What accounts are they using? Are they copying client information into their personal account, a business, an enterprise, a freebie? S- so many people are still on the free versions of these tools, yeah? So it's definitely time we started talking. You'll see a lot of people's note takers coming into meetings these days. You know, they're automatically transcribing. I don't see a lot of consent forms changing hands. I think it's really wild, actually, that I am very, very transparent and clear if you're working with me. This is being recorded. These are the reasons. This is the benefits. This is what I do. I use this particular type of AI at this level so that your stuff is more protected, et cetera. Um, you know, as much as I pay for all that extra stuff, I also use it when I'm private coaching. If I'm designing you an app, I'm going to use Fable 5, the king of Claude models, so that you can use some of my tokens to get a way better thing than if you're using your freebie, for example. So, um, it matters to me to have this higher level of security, but also higher level of use Browser extensions are such a danger zone here. Like, are they reading information? Are you logging into things? What happens when the project finishes? Like, at the end of your project, does anybody delete what was there? Like, okay, so that delete question that I asked before, imagine that the terms and conditions say it's retained for 90 days, but if you're at the end of the project, you've got everything you need in your CRM and final documentation, it is best practice to delete that conversation so within 90 days it then no longer hovers around the internet in the ethos, in the magic cloud. I, just side note, am so impressed with myself building... You know, being so ahead of this stuff and building all these GPTs and doing all of this stuff when I remember it took me, it's embarrass- an embarrassing long time to actually understand what the cloud was. Like, just being like, "But where's the cloud?" Like, "Yeah, okay, but where does it live? Like, what does it look like?" And I guess that's 'cause it was so long ago, but now I, it, it makes me giggle thinking of all of these things that I'm doing to be, and being ahead, especially at my age, and then how foreign the concept of just a cloud was to me. Anyway, let's get back. I'm making myself look like an idiot. Okay. So- What I'm describing, like not actually knowing what our third-party contractors and stuff are doing, is what some people call shadow AI, okay? So if you hear that term, people sometimes call it shadow AI, and it means that AI tools are being used within the business without the business necessarily knowing, approving, controlling that use. So it's like third-party use shadowing. And again, this isn't necessarily because somebody has done something terrible. If you are my contractor, hi Carly, if you've used AI, I'm not coming for you. I know you didn't do anything on purpose. It's often because nobody just created a policy or a rule. Nobody started the conversation. And I want to be so fair about this. Why would they have started the conversation? This is brand new to our brains. Like, these, these are all things that didn't exist even s- even six to 12 months ago when I was building Penelope. This wasn't a conversation that I was really diving into as much as I have done in the last sort of, yeah, six or nine months. So since building Penelope and thinking, "Okay, well, I'm selling this technology that auto-creates... I'm selling a technology that is an end-to-end system, will write proposals in like 10 minutes instead of a whole day," and it's beautiful. If you're interested, you can head to my website and check it out. But it's... And it's fabulous, and it still works. If you're interested, you can hit me up anytime. I'll pop a link in the show notes. But it is I'll pop a long-- I'll pop a link in the show notes. But it is a risk still, right? Because I'm asking you to put a full transcription of an in-home consult into ChatGPT specifically because Penelope is a GPT. Now, GPTs are locked down in a different way, and I'm not gonna get into all the backend stuff. So I'm-- I've done my research. I'm feeling good. But it brings me back to the AI clause that we talked about, because these are all things that are quite new conversations. If you downloaded my contract clause to start thinking about setting expectations for your clients around them feeding your AI work in, that clause is just the tip of the iceberg of the conversation that I actually wanna have. Because in the clause, it essentially helps you to communicate, "Here's what we promise." Your operations determine whether you can actually keep that promise, and that's where it becomes a business system conversation. I need a little button in this studio where I can hit, like, a little nerd alert button every time I talk about processes and systems. But they really, really matter, and they really matter to me because I see on the other side how good documenta- how good documented businesses connect with AI. Like, the girls coming out of Studio Build, it's a whole new world for them, and I love it. I also see, you know, just focusing on one thing. You're not gonna have to talk about contracts and clauses and internal policies for AI forever. You have to do it once, make some strong decisions, and sorry, like, big girl pants on. You are the founder of this business. You have to make the decision, write it down, draw a line in the sand, and then just review it every... I'd say probably every quarter because things are moving fast, so you just wanna make sure that you've got everything under control And we can't just say, uh, you know, you can't just say to your clients, "Oh yeah, I've thought about this, I've got it under control." Because if your contract says personally identifiable client information won't be uploaded into AI, who's gonna make sure that that happens? Is it just you? Like, maybe it is, and that's fine. But what counts as personally identifiable? Have you been careful to explain that? And does your contractor know? Have you gone and told your Carly? Does your VA know? You know, have you gone and told your Jess? What about meeting transcripts? Are they, like, just separate because you already get-- You know, do you get written consent when you go in? Do you get verbal consent and that consent is transcribed so you have it? Um, are you even asking that? I have always just assumed that people know when they're going into a house that you'd say, "Hey, I'm actually gonna pop this recording technology on now because I wanna capture every detail, and this is the reason why. Are you comfortable with that?" And actually getting people's consent. Consent matters. As a mum of three boys, I am s- seriously speaking about consent every single day, and it really, really is important. If somebody doesn't want you to do that, you gotta go old school. You've gotta get your notepad out, and you've gotta take really good notes in that in-home consult because they have every right to say no to having it transcribed. They have also got a right to say, "Yes, you can transcribe it, but no, I don't want you to put it into AI," and these are the, the things that we are gonna be talking about really soon What about putting in like a floor plan with the title block? Or what happens when someone, uh, like a- another team member, even yourself, you're sometimes guilty of this, is rushing at 2:45 to get to school pick-up on a Friday, they need AI to summarize something really quickly so that we could actually reply to the email before we go. Are we just gonna break our own policies, or like do we have an approval tool, a redaction process? Have you got a restricted data list, like things that absolutely cannot go into AI? Do we have a human review checkpoint? Like is there a process for who looks over this crap sometimes, slop I should say. Who looks o- over this AI slop coming out and supposedly going to be our caption for Tuesday? Like is there a human review checkpoint, because that is so important. Or is the entire AI policy currently stored inside of your head? Psych, there was never any policy. Don't come and pretend to me that you have a walking AI policy s- currently stored inside of your head. This is one of those examples where like the procedure doesn't live in the founder's head, it just doesn't exist, and that's so fine. There's no shade. This stuff is really new, and having a note in a CRM that just says, "Oh, Rhiannon knows what not to upload," that's not a business system. It's not sustainable and it's not, you know... I'm, I'm not a lawyer. I'm never gonna get into that conversation of what will and won't stand up in court. The legislation has not caught up with AI. The law hasn't caught up with AI. God only knows what courtrooms, you know, what courtroom conversations are going to be had in the future and what will and won't hold up. I'm not about that. I'm not talking to you from that lens. I'm talking to you as a founder of a, as a founder of a successful business, as somebody who works with design clients, who knows your workflows, your tech, what you're doing, what will help you, and how to get you to be able to access and unlock all of those AI wins without putting our head in the sand about this side of it as well. Because ethical AI is where I stand, and I fiercely believe in keeping creativity out of AI, not putting our, not putting anything that makes us amazing designers into these tools. They are great for so many things, but they are not great for that This is where I think the opportunity is. I don't want designers waiting for a client to send them an email saying, "Hey, can you please provide your policy on use- on the use of generative AI and client information?" And then suddenly you're Googling frantically or worse than that, you've just gone to Claude and said, "Can you write me a policy for the internal use of generative AI and client information?" Um, I would much rather in that scenario, you were able to go back and just say, "Look, absolutely. Here's how we use AI. Here's what we don't put into our AI tools. Here's how we protect identifiable information. Here's where the human review level sits. Um, here's how our team and contractors work within those roles. Here's our approach to third-party software that we use, and we, you know, we have someone that runs through that or we..." You know, you can take the terms and conditions of Canva, put it into Claude, ask for the key takeaways, ask it, ask it to look for any, um, anything that you should be alarmed about as a user who will be putting client stuff in. Like, you can do this work. It's not sexy and it's not fun, but if you're curious, if you're thinking, "I don't know the answer to any of the questions this chick is asking me today," then you can do a little bit of digging in that way. I shared in my series on Instagram this week, I'm doing one tip and one truth on AI every day. You can check them out in the highlights, um, if you wanna go back and, like, learn a, a couple of those tips and tricks and truths. A couple of those tips and truths. Uh, and one of them that I had shared is just how to, like, find all four keywords that will pretty much help you unpack the terms and conditions of your large language models. So that might be helpful. You can go back and check that out on Instagram. If you can reply like that, like this is actually our approach, that feels so different. It's not awkward, it's not defensive, it's totally professional. You're not gonna bother me. You're not like, "Sit down, let's talk about the data policy." But increasingly, I think this is gonna become a part of what a well-run modern studio looks like. So here are the questions I told you this episode is full of questions, and the answers will be addressed in my free webinar on the 24th of... The answers are something that I will be going into a lot more in my webinar that I'm running on the 24th August, but here are the questions I want you to leave... Here are the questions I wanna leave with. Don't try to answer them all right now. You probably won't know them. I just wanna... I just want you to listen, if you're walking the dog, if you're out there, going to school pick-up, you're driving, you're mopping the floors, however you listen to Designing Success, I want you to listen to these questions and just take note of your physicality. Like which ones make you squirm? This'll be fun. Okay. What AI-enabled tools currently touch client information in my studio? Which ones have I deliberately approved and which ones just snuck into my workflow? Like I've been using them for years, and now they've added additional AI, and I didn't even know. What client information am I currently uploading? How much of that information does AI actually need? What is necessary? Do I know whether my account is consumer, business, or enterprise? Do I understand what happens to uploaded files on that account? Do I know how long information is retained with my particular pricing tier? What happens when I delete something in my AI platform? Who else can use my workspace? Like who has access to my AI? What AI tools are my staff and my contractors using? What... Are client meetings being recorded and transcribed? Am I uploading intellectual property I don't actually own? That could be your photographer's images, like anything like that. Just think about that. Am I uploading intellectual property I myself don't actually own? What have I promised clients in my contract and privacy documentation? The privacy policy on my website, what does it say? Does it include AI? Probably not Do my actual workflows match those promises? And most importantly, if a client asks me about this today or tomorrow, this afternoon, next in home, can I confidently answer them This is exactly what I'm talking about in the webinar I'm running. It's a free session called What's Really Happening When You Upload Files to AI: The Truth About Client Data, Intellectual Property, and AI for Interior Designers. We're gonna get properly into this. We're gonna unpack it. I promise it won't just be questions. It will also be just a level of education around in the research that I've done, this is what I've found, this is what I think you need to know. It's not going to be a, like, scary legal chat about everything that you've done wrong. And when we know better, we do better. We're not going back into the past and just going to, like, stay awake all night every night being like, "I think I've un-uploaded this thing." What's done is done. You could go find that chat and delete it if you would like to. Um, you know, and then it may take 60, 90, 120 days, whatever your contract says with your platform. But you can only go forwards better, if that makes sense. Like, once we know this stuff, we can write an internal policy document. You can start to let people around you know what your expectations are. We're getting properly into it. We're going to understand the difference between processing, training, why your account type matters, what information you should think twice about sharing. Um, you know, what you should and shouldn't be sharing with AI. You know, floor plans, yes or no? Emails, yes or no? Marketing transcripts, yes or no? Client photographs, good, bad, yes, no? Supplier information, is that okay? Intellectual property that you own versus that you don't own. What should you be checking in the tools that you're already using? Like, where can you find the answers? And how to start creating some sensible boundaries inside your studio without becoming terrified of AI or producing 150-page policy that no one's ever gonna read. You didn't read the one from Claude or ChatGPT, so why are they gonna read the one from, um, designco.company or whatever your design brand is? I am absolutely not interested in terrifying everybody. Like, if you come along to the webinar, it's not a lecture. It's not going to... It's just education that I think we-- it's conversation that we all need to have. It's education piece that I think is helpful because confidence is everything and edu- like, just taking 45 minutes out of your day to be at a webinar or watch the replay. If you sign up to this webinar and you can't attend live, I will send you the replay. Just being able to do that can help you use AI in a way that feels ethical, feels confident, and feels like you've ticked your responsibility to learn a little bit more about the boring background s- um, stuff that you've signed up for. You may come out of that and think, "And I don't work for these guys." I say it all the time, but I don't work for Silicon Valley. I get no kickbacks. If you go from a pro plan up to a business or enterprise and you lock things down a lot tighter, that is a business decision that you have made. I don't work for them. I'm not getting a kickback. Um, I just want the businesses underneath AI, that are using AI, to be able to catch up, to feel confident, and most importantly, when you are on site, I want you to start Like start this conversation. Like if it comes from you, they're gonna ask you less questions. If you're confident with your clients before they're the ones starting the questions with you, and just have something that's like, "Yep, we use AI. Here's how we handle it responsibly." That is a way better position than being caught completely off guard when somebody says-- when somebody asks you these questions. Okay? So I'm gonna put the registration link in the show notes. Please come along or watch the replay. Bring all the AI questions you've currently... Bring all the AI questions that you're freaking out about, wondering whether you could... You know, who you're gonna ask. Come and ask me. I may not know the answers. This stuff moves so fast, faster than all of us, right? I'm not the expert, but I'm fully emerged in this world all the time through the lens of interior design and AI, so if anyone's going to know, it's likely gonna be me, and if it's not me, I'm likely gonna wanna find the answer for us both. I suspect we're gonna be talking about this stuff a lot more in the future, so we might as well just tick it off, get it done. I'm really excited. It's Monday the 24th of August. It's at 12:30 PM Australian Eastern Standard Time. The registration is in the show notes. You can also just contact me on Instagram, and I will send you a link. If you're on my email list, you will have seen it. There are hundreds of people already registered, and it's gonna be a really good conversation full of lots of different types of in- design firms and different ways of using AI. So I cannot wait Okay. Well, sorry if I freaked you out. If you have gone for your walk with the dog today and you're thinking, "Oh my God, I haven't thought of any of this stuff, and I'm just like a full AI cowboy," I'll see you in the webinar. Okay, bye for now. That wraps up another episode of Designing Success: From Study to Studio. Thanks for lending me your ears. Remember, progress over perfection is the key. If you found value in today's episode, go ahead and hit subscribe or share it with a friend. Your feedback means so much to me, and it helps me improve, but it also helps this podcast reach more emerging and evolving designers just like you. For your daily dose of design business tips, and to get a closer look at what goes on behind the scenes, follow @oleander_and_finch on Instagram. You'll find tons of resources available at www.oleanderandfinch.com to support you on your journey. Remember, this is your path, your vision, your future, and your business. Now let's get out there and start designing your success.