5 Ways AI Could Be Tripping Up Your Freelance Business - Bonus - with lawyer Ingrid Fernandez

AI is already in your freelance business, whether you invited it or not. 

The usual Being Freelance podcast is on its summer break, but here's a bonus treat for you: a replay of a live LinkedIn session Steve did with lawyer Ingrid Fernandez, part of the Being Freelance community, looking at what AI means for freelancers legally - not "will it take my job" but "how does it change my contracts, my responsibilities, and my relationships with clients?"

Ingrid walks through five areas that could be tripping freelancers up:

1. Putting client information into AI - what counts as personal information, why transparency with clients matters, and why your privacy policy needs to name the tools (including AI) that touch client data.
2. Clients handing you AI-generated materials - the IP risk when a client passes over AI-made brand elements, logos or source material without saying so.
3. Using AI in the services you deliver - the reputational and trust risk of client-facing work that's gone through AI without disclosure, and why it's worth putting in your contract.
4. Clients putting your deliverables into AI - what happens to your IP once a client feeds your design, copy, or podcast into their own AI tools to create derivatives.
5. Confidentiality clauses and AI - how modern confidentiality clauses are increasingly naming AI specifically, and why redacting a client's name doesn't necessarily get you off the hook.

Ingrid's steer throughout: this isn't about being anti-AI, it's about being transparent, thinking about where you stand, and then getting your contracts to reflect that - so you can stop worrying about it case by case.

Find Ingrid at herlegals.com.

Catch the full library of Q&A replays in the Being Freelance Community.

This was originally a LinkedIn Live - so come find us on LinkedIn:

Steve Folland on LinkedIn

Being Freelance on LinkedIn

Ingrid Fernandez on LinkedIn

JOIN THE COMMUNITY
You're not alone being freelance.
Come and hang out with your BFFs (Being Freelance Friends).
beingfreelance.com/community
 

NEW TO FREELANCING? THERE'S A COURSE FOR YOU!
The Being Freelance course is made for you!
Take the course and you'll also get 6 months FREE community membership.
 

FREELANCER MERCH
Get Being Freelance merchandise at beingfreelance.com/shop

Like VIDEO? - Check out the Being Freelance on:
Instagram - Instagram.com/beingfreelance
YouTube - YouTube.com/SteveFolland

Being Freelance is hosted and created by freelance podcast editor Steve Folland.

Transcript

Copy Transcript Copied

0:00

Hey, how you doing? I'm Steve Folland. Welcome to a bonus episode of the Being Freelance Podcast.'cause the usual one is on its summer break. But this is the live session that we did on LinkedIn recently with lawyer Ingrid Fernandez. I say lawyer like I don't know her. She's part of the being freelance community. This session looks at the impact of AI on our businesses, not as in taking our jobs, but as in how do our contracts, the way we think about it from a legal perspective, need to change. Whether we are using AI or not, it's hard to avoid it, right? So I think it's a really important one. I thought you'd like to listen to it. You can watch it as well on YouTube, Spotify, and also Apple Podcasts as you can with all of the past 18 months worth of podcasts for that matter. So, yes, enjoy. here you go. hadn't considered. Ingrid sent me an amazing list of five ways AI could be tripping us up. Everything from the way we might be delivering work to our clients, to maybe our clients giving at us AI stuff. But let's start at the very beginning 'cause it's a very good place to start. Are we putting our client's personal information in Ai? Do we need to define Ai?'cause AI seems to be in everything. It is in everything. It's in a lot of the tools that we use, probably without realising, it's sometimes on the back end of things that we might not know it's being used in that way. But the way I wanted to talk about it today is more when we're actively using Ai. So that might be Claude, Gemini, Chat GPT, Perplexity, any of the big LLMs that we see, we hear about. It's more, we could go deeper and I'm happy to chat about when the tools we use, use Ai, but this is more about the first gate we go through is if we are using it ourselves with our clients and the people we work with. Nice. All right then. So are we putting our clients' personal information in Ai? How, what could we be doing wrong then? So there are a couple of things that come up with this. Firstly is if we are doing it, are we allowed to do it? So that's when we have to go back to our trusty Privacy Policy, hopefully, which everyone obviously has and keeps updated every few months when they change their tools. But this is when it really does become relevant because we should be disclosing to people all the tools that touch their personal information. So for example, Steve, if you and I work together, it may be that Google Workspace touches your personal information because your email comes to there. And it may be that we have a Zoom call together. So Zoom touches your personal information and then if I was going to share some of the work that we do together with AI to brainstorm with it or to problem solve with it, I need to be transparent with you about what I'm sharing with AI about you. So that's the first layer, the transparency part. And then the second layer is there's obviously a lot of opaqueness opacity, o opacity. Uh, Hang on, I'll ask Claude no go on… There is that opacity around what happens within these LLMs and I don't pretend to know, I'm not an expert on how they intricately work, but when we are feeding personal information about the people we work with and the people who may be our clients or potential clients, we don't really have a lot of visibility over what's happening to that information. So we can put controls in place that limit whether those LLMs use that personal information to train itself. And we all should be doing that. But beyond that, there's just a lot of fog really around what happens. And this isn't to scare monger anyone, but it's that we just don't really understand how these things work and what they do with the data that they get. This is really about first transparency and then really making a discerning choice about whether you need to be sharing personal information that you're getting from clients or potential clients with AI and if there's any way you can do what you do for them without doing that. I like to think of personal information as something quite valuable that people are sharing with us in doing business together. And so when we are given something valuable, we have a responsibility to take care of it. So I think if I'm putting something valuable into this LLM that I don't really understand what it's doing, I don't know what's gonna happen to that valuable thing I've been trusted with. So it's really taking care of what we've been given. And when we're talking about personal information, that could be what. So your name, so Steve, and your email address, it can be personal details, like your address, your business address, your company business number. Then it can get a little more in, depending on the services that you do for someone. They could be the notes you take from the meetings that you have together. And that could have information about what kind of business they run, where they're at in their business, what challenges they're having in their business. You could have access to financial information about that company. It can really range a lot of the personal information that we come across in business might seem like business information when we are talking about it, but underneath it, we could identify someone's business, where they live, how they work, how well their business is doing, how much they're earning, a whole lot of information about them from that. So is the answer simply that this is hidden away in a privacy policy on our website? Or is it enough that I, at the beginning of that Zoom call or Google Meet call go, oh, do you mind if I take some notes, you know, or, Gemini, take some notes. Like, how should we be approaching it? I'm a boring lawyer, so I'm always going to say your privacy policy should have all these details hidden away in them. But on a more practical level, absolutely transparency should be involved in our conversations with our clients. So if you're having an AI note taker monitoring your calls, you should definitely just advise them and give them really the option to opt out. And this is something that we kind of talked about back in 2018 when GDPR came around and we were talking about personal information then and giving people a choice. And really we wanna be doing that again, it's not… again, I'm not at all demonising AI or saying people shouldn't use it, but I think we should be giving each other the choice of how we interact with it. And some people have a really strong position on that. Some people are really open to it. So I think when you get on a call and if you're going to have an AI note taker, give people the option to participate or not in that note taker. And if, for example, you use AI, and we'll come to this in a little bit in how you provide the services, let people know that it's, you know, it's one of the tools in your business and they might have a say in how their personal information goes into it. So it's that transparency piece if we can be open and make it a conversation, and I think that's the best way forward. Alright. Should we move on to question two? So question two was are our clients giving us materials that AI created? So this is something that is coming up a lot, especially for people, not solely, but especially for people in creative industries. So where they're taking materials, let's say for example, someone who works in web design or graphic design and they're taking elements to create a brand for a business and they might not realise that they're using AI generated elements that have come from the client. Now the issue there is very much an intellectual property one, and I don't wanna go down, legislation rabbit holes but it's really about what will you own? If you are giving someone AI generated material, which might not necessarily have any IP rights and they're creating something from that, will you actually own your logo and would you be able to trademark that? So these are really practical implications that are coming into our businesses and how we work without properly being aware of them. If we jump onto Chat GPT or Claude and we get it to mock up a logo for us or a design of our, our brand name, and we then pass that to a designer and say, we want this particular font and we want these particular design elements, not realising that we might not then be able to protect it going forward. So I think again, a lot of this is gonna circle back to that transparency. It's not again, about demonising the use of Ai, but it's being open about the people we work with. So they then know how they can use the materials and, and make sure that we are protected. So we could be asking our clients, was this material created by Ai or just so I know, just so I can provide you with the best thing, that kind of statement. Yeah, absolutely. So a lot of the people that I'm working with, we are putting that into their contract that clients, when they share materials with us or source documents with us, to just let us know that some of it may be AI generated so that we can then do our work with all the knowledge rather than we doing it thinking we are dealing with client information and we're actually dealing with AI generated information. Which actually feels like,'cause sometimes a contract, I, I know you, I know you are a lawyer, so a contract is important, but this actually feels like really important stuff for a contract, especially if you are a designer. Like the example you just gave where, you know, like they could come unstuck further down the line and then pin it on you and you can say no, 'cause actually I'm pinning it on you because it was there to begin with. Hopefully we work nicer than that and we can say we had a discussion about this at the beginning, but it is in 4.2 that Ingrid made for me. That's exactly, it's obviously avoiding any arguments or disagreements, but it is, again, drawing our client's attention to things. And it's because a lot of us aren't thinking about this yet. Again, it's not to make it seem like people are trying to do anything sneaky, it's nothing to do with that. It's really that we are using this tool so much in our businesses and again, I don't expect anyone to think about the legalities because they're not the most exciting part, but they really do have legal impacts for our businesses. So it's about having honest conversations with our clients and having it in our contracts so that we're not left liable just in case something goes wrong. Okay. So number three is, are we using AI in our services for our clients? So this is an important one because a lot of us might be using AI in our businesses in different ways and not really realising what that implication is for our clients and our customers. So for example, a lot of us might be using it for ideating, brainstorming, as kind of like a sparring partner. And again, I'm not saying any of this is bad or good or wrong or right. It's just about recognising where you're using it and what you're using it for. And then noticing if any of the deliverables that you provide to your client or the end services that you provide to them have been created without your input. And whether it's just an AI agent generating that output. This is a really important consideration in terms of what the client is getting, because one, obviously we know that AI makes mistakes and it hallucinates and it makes things up. So from a factual perspective, it's really problematic, but it's also because we can't always see how AI gets to those answers. We don't know the process it went through. It might be relying on the information that we didn't give it, that it's sourced from somewhere else. So one of the reputational risks really is that we don't know where these outputs are coming from. We don't know where these deliverables are coming from. So where we're trusting AI to do parts of the work that we would usually do, we just run the risk of it not being something that we would put out into the world or that we would accept money for to do for someone. So it's noticing, I think, where AI can be a really valuable tool for you and your business, and balancing that against the risk that it runs. If you know, I can understand that when people get really busy and things are very hectic, it might seem easier to trust AI to do part of your work for you. But if that's something that's client facing, we really need to consider the risks that's going to run for us and for the client as well. So this is more a thing for us to mull over, or is it also something that should be in our contracts? I suppose it's back to that transparency again as to saying, Hey, in some instances I use this and maybe explaining why, 'cause it makes it more efficient so that I can provide X, Y, Z, but I don't know what's, what should we be doing? Absolutely have it in your contract. One, it's the transparency that, you know, keeps popping up as we chat but another part is to protect you. Because if people have that transparency about how you're using AI in relation to the services and products that you provide to them, they're more likely to engage with you honestly, knowing that you've used it as a brainstorming tool or you've used it to synthesise a lot of information so that you can then provide them with the deliverables. Whereas when it's hidden, I think especially now where we are in, you know, what people are calling a low trust era, if we are getting things from people that perhaps sound different to how they sound when you talk to them and you might think, oh, that doesn't seem like the Steve that I had a chat with, or that doesn't seem like Steve on the podcast. This is a really different output. It really breaks that trust with our customers and clients. So if we can be honest and open with them and say, yeah, I, I use this for some part of the services, but everything I give you at the end of the day has my fingerprints all over it. Everything goes through me before it comes to you. It just reestablishes that trust and having that in the contract gives the client the peace of mind that it's you that actually engaging rather than Claude. Yeah. You see, this is why I like to miss deadlines sometimes, Ingrid, because it just, it's that human element. Do you know? It just makes you so much more human to everyone when you are not where you're meant to be. I'm joking. Of course. I love to hit a deadline. Okay, we're working our way through five things that Ingrid was trying to get us to think about in the way we currently run our freelance businesses, the way we're incorporating AI into it and how that could be tripping us up. And the fourth one is, are our clients putting our deliverables into Ai? So this is the one that I think would probably be front of most people's minds. This is what is occurring to most of it. So the other things are things maybe we need to think about, but I know this one is, is worrying a lot of people. So if we go back to the designer example I gave before, if a designer is producing a brand and a logo and beautiful design elements for a client and then the client feeds that into AI to make derivatives of that, that they can then use on social media or that they can then use on a flyer. There is so many unintended implications, one for the designer's business, and obviously we're all aware that, AI may be taking people's work and people's jobs, but the other part of it is the real creativity element. At the end of the day, we are humans working with humans and humans trusting humans, you know, in these transactions that we have for our businesses. And I think when our work that we put our effort and our expertise into is taken and put into something where other things can be spat out without the human part of it, I think that really risks not just our jobs and our work, but the human relationships and interactions we have by doing business with each other. Because could you just have a robot creating the design element and then sending it to a person who puts it into their robot, who then sends feedback to the first robot? You know, at what point do we stop having humans in the loop at all? If we are just going to use these robots for feedback and for interactions and to ask questions and to draught emails, it's not, again, I'm, I sound a bit anti AI and I'm not anti using it, but I am anti removing the human from the loop in And how can this trip us up legally? You know, there's the moral side of it. I mean, does it come down to the contract again? So if I'm an illustrator and I've created some characters for a client, but then they've taken those characters and they've fed them into an AI video generator and started creating a video yeah, is that where this becomes an issue? Because I could have, maybe as the illustrator, I could have animated those characters, or you are now, it's a bit like if you were to put those characters, like in the past, the, the contract might have said, you can't put these on a t-shirt and sell them, for example, without paying me extra, uh, that sort of licence. So is that where we're looking at issues here? Yeah, absolutely. Thanks for bringing me back to the legal. I go off on my ethical rants, but it is definitely, it's definitely a legal issue because it's our intellectual property. What we create, we either own the rights to, or we assign the rights to, or we licence the rights to whatever we choose to do. We still own the rights. And as soon as we hand those rights to someone who puts them into a robot, we then lose control of what happens to our work. And so when these robots are then creating derivative works, or creating alternative works, one that we could have used, or two, using them in ways that perhaps are not covered by our contract and are not part of the agreement we made with our client, that, again, puts our businesses at risks. Because if our designs are out in the world for things that we would not have agreed to, that again reflects badly on us and our businesses. That might not have been a client, you would've signed up if you understood what they were going to use those designs for or in those particular ways. So it definitely has a very real legal implication. And yes, your contract is the first point of call because we, this is really where we can have our biggest protections in place, is in our contracts. And We wanna have relationships and conversations, but our contract is where it sits. If something goes wrong, we wanna have a mechanism where we can take action if something happens. We don't just wanna be stuck with the bad feelings or the resentment. We wanna be able to do something. And that's where having this in your contract and restricting people from using your work in that way, that's where it can give a business the best protection. And I wonder whether, is there a blurred line where, for example, a designer could give branding guidelines and everything to a client, and then that client puts it into Canva and then Canva Ai and or could be other AI is available, uh, all doing the same thing could be Adobe, right? But the my thinking is, is that, is that different to them giving the brand guidelines to a different designer and paying a different designer to go off? As in where does the job finish? You've created the brand guidelines, now the client can go away and create whatever they want with those brand guidelines. Or is it simply about defining that in the relationship, in the contract, Definitely it depends on where you land. So there are going to be some people who want to restrict what people can do going forward, and your contract can do that for you. And there are going to be other people who just worry about what happens in that relationship. And then once the relationship's done, they're happy for you to go do whatever you like. And that can be in your contract. So this is the benefit of having a contract that works for your business because you can make it do what you want. Obviously we're working within the law, but when it comes to your IP you can be as strict or as flexible as you want to be, but if you are not thinking about it, and if your contract is silent on it, then you can't do anything. You relinquish your right. So it's really about thinking about this and then putting it into your contract in a way that gives you the powers that you want to have. And no more than that. Just as far as you wanna take it. Yes. Yeah, because I'm think, I'm a freelance podcast editor and video editor, so I make a lot of video podcasts for people, and I always kind of feel like, well, you've hired me to make that video podcast, I give you that video, and then I don't really care if you want to feed it into your own AI thing of choice to make the clips, to, to sort of cut things up. Even though you could pay me to do that, it's your, you know, you can go do that. It won't be as good as if I did it, but you can go do that. But should I maybe be warning my client that, they're, they're about to put the guest of that podcast into an AI thing. Like Yes. Even if it's just showing them what a professional I am, even if it, they then go and do it anyway. Yeah. Again, we can't, you know, as much as we can have contracts in place to minimise what people do with our work, we can't control people in their business and what they're going to do. So, like you say, they might take the information you've given them and then go do what they're going to do anyway. But it comes back to that transparency. So if you know that they're going to do that, and if you communicate that to the podcast guest, they might have feelings about it. So again, it's just giving everyone that say, having the documents in place so that if you want to take a particular position, your contract supports you doing that. But then using that as a springboard to have the conversations to be transparent, recognising that people are going to do what they want to do anyway. But here are the legal implications that you are trying to minimise for your business. Okay. Our fifth one was, do we have confidentiality obligations? So basically confidentiality obligations that restrict what we can put into Ai? So they can, depending on what your contract says and the different contracts you have with different people, so your client contracts might say something, if you work with other suppliers, they may say something else. So this is a good reminder to go have a look at your contract. But the confidentiality clauses in a lot of contracts that we are seeing now have specific limitations on sharing any business information that's considered confidential that comes up during your client relationship, restrictions from putting any of that into Ai. So we used to have quite standard confidentiality clauses, obviously some were a bit more bespoke than others, but they tended to say the same things. But we are really seeing now a lot more confidentiality clauses are being strict about AI specifically. So if you were to enter into a contract with someone and didn't really check the confidentiality clause, and then you put some of the materials they gave you into AI to do some brainstorming or to do some of the background thinking work, you might be in breach of your contract and then you know there are legal repercussions you could be facing. So it's just another area of the contracts that one we need to look at in terms of the contracts we are signing. And then in terms of the contracts we are giving out to our clients, do we want to restrict our business information being put into Ai? Do we want the confidentiality clauses to name that AI shouldn't see any of our business information, so we need to think again about what lines in the send we wanna draw in our contracts. Flipping heck. What if I had client information and I redacted the client's name from it, and then brainstormed using, Chat GP Claude. So it's going to depend on what the confidentiality clause says. So their name is personal information, but if it's, let's say, information about their business. So let's say they've come to you to produce their podcast and they send you some information about their current financial situation in terms of sponsorship that they're getting, or in terms of the amount of listeners they have per month or whatever it may be. And you then feed that into AI that for them might be considered, or according to the contract, that might be confidential information. So even though you've redacted their name, the body of the information might still be confidential and you might be breaching what the contract says by feeding that into ai, even if it's just for you to be able to think and process and brainstorm. So it's not necessarily just the personal information, it's also what else we are sharing. What if though, that comes at a stage before a contract? Because sometimes you have discussions with clients and you are sending stuff back and forth, brainstorming ideas. You might create a pitch for something like, in some instances you might be asked to sign an NDA, but not always. No, absolutely. There are a lot of times we are dealing with confidential information without having that contract in place. And what a rule of thumb that I tend to give people is deal with the confidential information you get, how you would deal with your own confidential information. So it's really minimising the risk for any confidential information or potentially confidential information. So anything about anyone else's business that you may come across or that comes into your possession or that you work with, treat it as confidential as you would treat your confidential information. I'm imagining I'm watching this and thinking, well, a bit like Keri has just written on LinkedIn, Hey Kerry, she's written, feels like a minefield. It's a polite way of, of saying, what I was thinking is how do we go about this? Because we are companies of one, so we need to think about, it feels like we need to sit and think about how we feel about this, but then also think about how we want this in, in, in our contracts, what we might get from client contracts. Like what's the step, how do we navigate that minefield, as it were. Yeah, absolutely. And I'm, I'm so sorry, Keri, this isn't meant to be overwhelming, but it is a little bit of a minefield because we are dealing with this new tool, and when I say new, I know it's been around a few years, but it's new to the business space and we are seeing it in our businesses in ways that, you know, the law wasn't necessarily designed for. So we're, we're kind of working in a way that we think is best with the laws that we have. So the first step is definitely to have a think before… we'll come to the action step in a second, but I'd really get people to just think about this stuff. How you feel about your information going into Ai. If it doesn't bother you, great. If it does bother you, then you need to think about what that means for your contracts. And then think about how you use AI in your business. And again, there's no right or wrong here, but how you're using it in the best ways, how your clients may feel about the way you use it, how transparent are you being and how you use it. And then once you have your head around what you want that to look like, you can definitely get legal advice about protecting yourself and protecting your business, protecting yourself in your contracts because the first step is really to understand where you stand and then get your legal documents to support that, because they're really there to support your business. They're not there to be a blocker or to make things more complex. It's the opposite. Your contract should actually make all of this more simple because then you're not thinking about it each time. It's just there on the piece of paper. You send it out and you're done. You know, you don't have to worry about it each instance. Such a good point. Okay. So we can have a think about it, look at our contracts, talk to a lovely lawyer like Ingrid, who can, who probably, let's be honest, is more than us thinking and seeing lots of these different scenarios and therefore having to update stuff. It feels as well, like some of this is probably like, I'm not demeaning your job, Ingrid. No, It's probably not adding another sentence onto an existing thing. It's like this clause and, the Ai, I'm not saying I could write that sentence, but it feels like a lot of this is already there, but actually we're expanding the way our contract might already exist. That's exactly, so this is not about tearing everything down and starting from scratch. It's really just refining things because of the tech that we are dealing with, protecting your business a little bit better because of the tech we're dealing with. So it's not at all an overhaul. It's really just tweaking what you already have to make sure you're as protected as you can be. You're as transparent as you wanna be, and you're doing the best you can. That's, that's really what it's about. Nice. We'll put a replay of this in the Being Freelance Community. If you dunno what that is, it's some, it's full of lovely freelancers. Ingrid's one of them. You can go to being freelance.com, click on community, and uh, you'll find us all in there. And it's got replays for q and as and lots of other things that we do as well. So, the main things that we were talking about, was starting with the privacy policy, which is most likely, hopefully on your website already, which we should be updating, I think you said, with any new tools that we use anyway. So in this instance, the AI elements of those tools. So that doesn't necessarily mean we have to then outline Google Meet's own terms and what they might do with it. It's simply saying, I use this tool. That's right. Yeah. So yeah, privacy and then moving on to thinking about how we want to be transparent about it and therefore hoping our clients will be transparent with us about our use of ai. That's not really a legal thing. More just a sort of moral and doing nice businesswise. And then in our contracts, how it will affect us doing business, but what our clients might also do with the work we create. So all of the things that probably already in your contract, do they need to be expanded, which is where a lawyer can come in handy to go, yes. And then put in the right wording. You are not just protected, but it just makes things simpler as well. Once it's in there, you can stop thinking about it and worrying about it. Yeah, absolutely. It's just one of those things we want to sort as much as we can so we can get on with running our businesses and not have to think about it. Brilliant. Ingrid, thank you so much. See you soon. Everybody else, thank you for joining us, Sooooooeerz as sooz? As deal with whatever that word was, transcription tools. I dare you even, I don't know what I said. See you soon guys. Take care. Bye Bye. Thank you very much to the brilliantly, lovely, wise, and generous Ingrid Fernandez. You can find her at herlegals.com, herlegals.com. I'll put a link of course in the show notes and, uh, yes, as I mentioned, do come find us in the Being Freelance Community. There's loads of replays for loads of great Q and As that we've done over the years. Okay, I'll see you for another one. Well, not that soon, 'cause this is meant to be a summer break, but who knows, maybe there'll be another bonus. Keep checking. And there's loads of episodes from across the past 11 years for you to catch up on after So yes, see you for another one as soon as we can. In the meantime, you have a great week being freelance.