The AI Mistakes That Could Destroy Your Channel

[00:01] business owners are using AI for thumbnails, scripts, and logos, but most don't realize that they could lose ownership of their own content or even get sued for what AI generates. >> If you were just typing in a prompt and

[00:16] taking what the AI gives you, that is not protected by copyright law. You basically have no ownership of that. >> Today, we're joined by Autumn Witt Boyd, who's a trusted attorney of seven and eight-figure online business owners and

[00:29] their intellectual property, their contracts, and their revenue at scale. >> You could accidentally copy someone else's work because, again, AI is pulling from its library of things that have already existed.

[00:44] >> to get the playbook that most business owners wish they had after making expensive mistakes. What is the single biggest legal mistake that creators and online business owners are making when it comes to AI right

[01:00] now? Yeah, I think the biggest thing that we see is they are relying on it as if it's Google and it's giving them the right answer. When actually, AI is generative. It is pulling from past. It is searching the all the language that

[01:13] it has taken in and it's giving you a prediction of what it thinks the right answer might be. But, if a law has changed, if you're throwing a, you know, a new scenario at it that it is not in its library, it is just going to make

[01:27] things up. It's going to hallucinate. And so, relying on it for, you know, answers to compliance with laws or what should I put in my contract can be just totally wrong. >> Is there any examples of like stories

[01:40] that have happened of someone relying on it and then it backfiring? Yeah, we see this a lot with our clients who have big teams. So, the employment every state and they change every year. And so, you can say, you know, "What's

[01:54] the rule in Arkansas? Can I have a non-compete in my contract?" And it for that." And it could be totally making it up, or it could be out of date. So, that is an area we're seeing a lot. So, there's like a couple big areas

[02:08] I want to talk about, and we could go through like different areas of legal risk, like ownership and copyright, data privacy, client risk, or contracts and team. But let's start about ownership, because a lot of online business owners

[02:22] and creators are publishing content, of course. If I use AI to write my script, do I own it? Maybe. This is a lawyer's favorite answer. It's the worst one. So, we're recording in February 2026. I like to say that cuz this is changing

[02:36] really fast. But right now, what the US Copyright Office is saying, and this may countries, but what it is saying is that if you are using AI to create something, if there's no human really input, if you are just typing in a prompt and taking

[02:53] what the AI gives you, and you know, using it as a script for your video, or even creating, you know, video content, music, things that you're incorporating into your videos, that that is not protected by copyright law. So, you

[03:06] basically have no ownership of that. What that case There's a case that's working its way up to the Supreme Court that's challenging that, so stay tuned. That may change. The bigger open question, which is what I think is more

[03:19] relevant to your listeners and viewers, is what if I use AI to help? What if I kind of give it some ideas, maybe I give it an outline, and then I add my special it an outline, and then I add my special sauce? How much AI is okay, you know, to

[03:33] get protected? And we just don't know, but our recommendation right now is the more human, the better. So, I like to say use it for brainstorming, for helping you come up with ideas, maybe even that first outline, but then really

[03:47] putting your human touch on all the stuff in the middle. And then you could use it again at the end, say, you know, give this a proofread or are there make sense, you know, cuz as kind of that last review. But using it to create

[04:01] a script that you just read is a bit of a dicey area. Can you copyright AI answered that, but I'm curious. >> right now, no, you cannot because it's considered machine created and our copyright laws require that human touch.

[04:17] I saw a story from Suno, which is blowing up, and that is where people are making music and they're making really original songs, but it's completely made with AI based off prompts. Are you familiar with that? Is that And that's

[04:29] those songs. Yeah, so there is an argument being made that there is creativity in the prompting. And you know, if you get one result and then you like the case that's working its way to the Supreme Court, there's like hundreds

[04:44] prompting. Um, right now the copyright office is saying, no, that that's still is not protectable. That may change, we'll see. So, I guess then the question becomes is us asking about risk because, okay,

[05:01] we have uh content creators that might be make making AI songs, YouTube background music, podcast intro music, um, whole faceless channels where maybe you're doing devotional or scripture content over music, meditation content.

[05:17] happening. It Maybe the question isn't can I copyright it? Can I monetize it though? And it would seem that in some cases you can. However, as we speak, there's mass demonetizations happening on YouTube

[05:30] right now of um, what they would call like generic AI um, what they would call like generic AI slop or like generic repetitive content. And so, I If there's the copyright question, which you answered, there's

[05:44] the can I use it though question. Right. >> And then can I monetize it? Can you break that down? Yeah, so the can I use it is going to depend on YouTube's terms of service. Like that's their decision as a platform, what do we want to allow

[05:56] here? And that probably will continue to evolve as hopefully the output gets better and more interesting and it's not just AI slop. Um but, you know, YouTube has a responsibility to the users and the

[06:08] content creators to keep it a high-value place. So, they're going to want to cut Um now, can I monetize it? That's also a YouTube question, but I think the more unique and creative that you can make your content, the more likely

[06:23] it is that YouTube is going to allow monetization. Um the other risk here is that you could accidentally copy someone else's work because, again, AI is pulling from its library of things that have already existed. So, it's not

[06:36] creating it out of thin air. Um and so, we've seen that with images. It's especially risky if you're trying to create, you know, a brand or a logo or trademark. You could accidentally step on someone else's toes without even

[06:49] realizing it cuz the AI is, you know, pulling from that library. And then what would that What could that lead to? Being sued by them, them pulling up and telling you to take something down? Yeah. Yeah. If you're an entrepreneur or

[07:01] online business, that's why we created the Think Media Mastermind. >> I have so much more clarity as to my ideal target audience now, which means my content is about to be so much better

[07:13] person I'm trying to reach. >> Super intimate, high-level strategy. I had the skills that I already knew sharpened. I feel like I went to my next

[07:25] level. For entrepreneurs and creators that want to scale with YouTube. This was the first time that I was able to get in a room with a lot of other serious YouTubers and talk with other people who love creating content and

[07:38] love YouTube. Usually I I get to do that. So, this is really special. You thinkmediamastermind.com. What happens if somebody copies my What happens if somebody copies my content and uses AI to copy my course or

[07:51] AI to copy one of my videos? Yeah, so that's copyright infringement right now. There's again a lot of litigation happening here because the AI libraries were built by taking other people's work without their permission. And so they're

[08:03] different way because we're using it to train the model and we were doing it's fair use, which under copyright law is okay. in the scenario you're talking about, I'm imagining you think of a competitor

[08:19] content and, you know, using it to create basically competing content. That's always going to be copyright infringement. That's never okay. I mean, think that's frustrating, but how would you manage that or how would you go

[08:34] after that and is it even worth it? >> Yeah, so there are robots that you can train. Now, they're all AI enabled. Uh, to kind of crawl the internet and look for things that are similar to your works. So, if, you know, there are um,

[08:48] then there are services that will send takedowns and cease and desist letters and you could eventually file a lawsuit. You've got to manage where you're putting your energy. So, I often say, you know, we want to protect our IP. We

[09:02] and that our content is our own, but also it's like whack-a-mole sometimes. profile gets, uh, the more you will have people copying you and most of them are making no impact at all on your sales, on your

[09:18] followers, you know, no one's confused about whether it's Sean's channel or, you know, Joe Bob's. So, we we kind of take those one by one. They, you know, if you do see someone who maybe used to work for you and is using your content

[09:31] and has kind of an inside view, we want to shut that down for sure. Um, but sauce and people follow you and learn you teach, the background you bring to

[09:45] long way. You know, you've got this framework that I want to cover in these before we get there I want to go deeper and try to say, "Okay, but what do we even do about this besides worry and

[10:00] experience anxiety?" You know, that sometimes a topic like this is it just feels overwhelming. There's so much AI and ignoring it or putting our head in the sand might not be wise, but for so many listening to this, we're like,

[10:13] "We're really not that big. Who's going to come after me? Should I really even worry about this? If I use AI to generate a a thumbnail or pull from some things, am I even going to show up?" Like how do you have a decision-making

[10:28] framework to move forward so you don't get paralyzed by analysis paralysis and anxiety? Yeah, so my perspective, I'm one lawyer and I forgot my standard disclaimer. This is not legal advice. So I'll say it now.

[10:41] Um I like to look at how much do we need to worry about legal as kind of a spectrum and it's depending on how much risk there is in your business. So if you're just getting started, your risk is very low is the good news because you

[10:53] don't have much revenue. No one's counting on this. You don't have a huge team. Um so you need to be careful and you need to be building. You need to be putting out content. You need to be um

[11:06] you know, growing your channel, growing your revenue. Um as you grow though, as you have more eyeballs on you, as you have more more at risk, now you've got clients. Now you've got maybe a team. Your family is relying on your revenue.

[11:20] Now we want to start to build in more protections. So the it kind of the the legal protections and the amount of worry kind of scale as news is as your business scales, you have more resources. It's not just you

[11:35] the things. You can pull in, you know, whether that's buying a contract template if you want to make sure your website has the right policies or hiring You know, you you have the ability to get some help. So, again, we're going to

[11:48] go on to the second category, but I would love for you to talk about this. would love for you to talk about this. So, the New York Times is suing OpenAI and Microsoft for copyright infringement alleging that the companies used

[12:00] authorization to train AI models like ChatGPT. That was filed in late 2023, but it continues to go on and it's claiming that OpenAI's models can generate near verbatim paid content by bypassing the Times paywall and

[12:16] threatening their subscription revenue. What do you know about this case and what does it mean? Yeah, so it's ongoing, excuse me, in the courts. forth. It is an it is still an open question. So, this is what I mentioned

[12:30] earlier how there are lawsuits where the companies that are training the AI on service. They they did all kinds of things. This is the worst way. If I were in-house counsel, it's mortifying. They clearly decided that this was a

[12:45] risk they were willing to take. So, it's an open question and I think it will determine a lot of the future of how these models continue to grow because they are scraping from the internet all of your content, my content, um, often

[13:00] really a way to stop it is the hard part. There's also some authors that are suing because their books were used to train the AI. Um, they're not happy about that. Um, I think it's a honestly, I think it's a

[13:13] pretty clear-cut copyright infringement, but um, and it basically says if you're using someone else's work and you're doing something interesting and new with it, that's okay. So, these are things like

[13:26] parody or commentary. And so that's the argument that the the AI companies are something different. We're making this new tool that's going to revolutionize the world. And so it should be fine that we didn't pay for any of the materials

[13:40] there there's no way they could have. The The The New York Times wouldn't have licensed it. So they kind of decided ask for forgiveness rather than permission. Um it's an open question what the consequences will be. I can almost

[13:53] than any damages they will ever have to pay. So um it it kind of feels a little bit like we're chasing like the dragon is out of the cage and we're chasing after it. Uh even if the content owners win, I

[14:06] victory. >> If we're still talking about ownership and copyright risk, one of the questions that creators hesitate from and and they want to make content, they maybe want to jump

[14:19] into cultural commentary and news, and you think about this idea of using a president in your thumbnail or using Darth Vader in your thumbnail and using content in your video. How should creators be thinking about navigating

[14:35] creators be thinking about navigating likenesses of real humans, IP from uh there's also been Disney and Universal sued Midjourney over images. Universal sued Midjourney over images. So that's the AI side of things, but a

[14:50] lot of individuals are like, "Okay, I could I could get some views off of jumping on a trending topics, but what How do I stay safe as far as using these images in my videos?" Yeah. So I don't know what the YouTube takedown policies

[15:05] that you're probably you probably know more about than I do. From the legal perspective, it's copyright infringement if you're using someone else's image or photo or music without their permission. Like full stop, there is generally not a

[15:17] defense. What you've described using a picture in a thumbnail um is usually going to be infringement. Now, if it truly is commentary, like you're talking about the president's policies and you flash a photo of him,

[15:29] that might be okay. Um, or if you're doing a parody, you know, something funny about something he did, uh, where you're kind of making a statement with it though, you're not just using his photo and making a joke.

[15:41] Uh, but it's a very gray area. So, that's what this this case with, um, the New York Times versus Open AI, it's a new theory, like they're going to be making new laws. So, they have to go to court. It's not like a defense that will

[15:55] stop someone from suing you. Like if you are relying on fair use, you are going to court and say, "No, it fits in this category, so it's okay." So, I always recommend, if you can, to get a license. And some of these things, like Darth

[16:09] characters, are harder to get a license for because they want to control that a little bit more. But, a lot of things you can just find a picture that's, you know, a dollar 50 on Shutterstock or something. Um, music, the same way. It's

[16:22] not super expensive, generally. I would say the thumbnail is probably the bigger problem than if you're using it in your video as kind of part of the point you're trying to make. Yeah, so, you know, that makes sense and

[16:34] I think you would drive us to be, uh, conservative on the video version. Uh, on the video version of the podcast, um, we're going to bring up this image that I think, um, will be very valuable for individuals watching. Um, fair use

[16:51] has been described as this, that if you only use a small portion of the material, and if you're commentating on the new Star Wars park at Disney, or you went to Disney, then Darth Vader's in the thumbnail. If you're giving your

[17:04] take on the latest presidential policies, if only a small portion of the original material is used, if it's a new work, or like you said, critique or satire, uh, or education, then if it also

[17:16] benefits the user and is predominantly other than area cuz is it commercial if you're making money from it? >> Right. And then the new work is predominantly an original product, not

[17:28] trailer and re-uploading it, but you're commentating on things. And the isn't diminished. It also in some most cases probably amplified, positive or negative. It's more press and media. You know, a lot of gaming companies are

[17:43] and put their intellectual property in the thumbnails. What's your take when you look at this fair use chart? I feel like you got an A+. Sean, that is great. >> So that all each other you can if you watch the video, it's worth it. Um

[17:57] each of these categories that you just went through and those are basically the The one thing you didn't talk about, but it's kind of included is this idea of transformation, which is talked about a lot by the courts. Are you transforming

[18:11] like you said, you're not just showing the trailer as a trailer, but you're subtracting something, making it something different. That has become a really important thing that people are looking at. Brilliant. So definitely um

[18:25] podcast listeners on audio, but you definitely want to visit the video version just to see that and that can help you uh have a level of confidence you mean I I think I know what you're going to say, but it's like

[18:39] the idea of everyone's doing it and getting away with it. Yeah. So it must be okay. How do you feel about that? Well, I think that is important. Like then there is figuring out what your risk tolerance is and maybe I'm not

[18:52] following it 100%, but I I understand the risk and I'm okay with it. Um if everyone else is doing it, that's probably a good indication that maybe the copyright holder, like you said, like the video companies who

[19:04] their games cuz it's like free marketing. You know, there are some things that are probably technically infringement, but Um now, you can still get burned because

[19:16] they can decide, "Well, you're the one that we you know, we don't like." Um but worst thing that happens is your video gets taken down. Like maybe that's an acceptable risk. >> Yeah, that's good. Okay, so that first

[19:28] big category we're talking about like the three biggest legal risks of AI. We're kind of expanding into other areas on YouTube. The first area was ownership and copyright risk. The second one's data privacy and client risk. So, if I

[19:43] upload client information into AI tools, is that okay to do? It depends. So, does your client know that you're uploading their data? I actually I went to a legal exact topic, so it's very fresh. Um we're adding it to all of our

[19:59] contracts with our clients, just giving people notice of how we're using AI, so allowing them to opt out. It's basically if you want to work with our firm, this is how we do business. Um because it's too difficult to kind of figure that out

[20:13] person by person. Uh but I do think and there may be depending on your requirements that you give notice to your clients. If you the problem is with security and um the free versions of a

[20:26] lot of tools are training on whatever data you give them. Um some of the paid you give them. So, you have to really be careful with those settings and make sure uh especially like a lawyer, a CPA, you know, some a doctor, you're dealing

[20:40] with uh very sensitive information, you want to make sure that you're honoring requirements. You're not giving away confidential information that should not >> Okay. And then what happens if there is a data breach?

[20:53] Yeah, it varies by state, but there's all kinds of rules about how you have to tell people, send out notices, potentially, you know, buy the uh and the security system for them to have so they can watch their

[21:06] uh credit score for 3 years depending on what kind of information was let loose. So it's embarrassing, it is expensive, it is not something you want to be dealing with. >> And it's embarrassing and expensive for

[21:20] about are you talking about the individual user? of the scenarios we were talking about last night were let's say you have an employee on your team who's using an AI tool, maybe you

[21:34] they're they're they're uploading sensitive client information that then there's a data breach and it gets shared. There's no like clawing that back. Yeah. >> Um so the you know, you have a risk just

[21:48] the vendors you're using and a lot of these AI companies are really new and it's exciting, they're doing really fun things, but you know, they don't always put security first because they're innovating really fast. And when you

[22:02] also think about like you know, we're getting more technical, but even if good information to have as they think about scaling their online business, their content creator business. But what is GDPR or what is things like

[22:16] specific laws in California, just you know, consumed or where your clients or customers might be? Yeah, so GDPR is the law. They give their citizens and residents a

[22:32] lot of data privacy rights that we do not enjoy here in the United States. So countries, you are going to want to pay attention to that. It's mostly most of the rules around how you can use people's data and that could

[22:45] them opting into your newsletter or downloading a freebie. It could be their campaign, you know, you're sending them updates or reminders. security number or their bank information. It can be something pretty

[22:59] requirements about how you can handle that information, whether you can sell it, and how you have to tell people what you're going to do with it. California does have some privacy laws. They mostly apply to larger companies. So most most

[23:12] of our viewers they're I think it's 20 million is the annual revenue floor for that to apply or 100,000 customers in California. So it's a pretty that they have a pretty high bar. So definitely need to be aware of that

[23:26] and if you are plugging your plugging this data into AI, that's just one more thing that you have to be careful about if people ask you, you know, to to take them off your list. >> Do you have a particular favorite LLM or

[23:40] workflow you like to use because of the level of security? Yeah, well we use a legal one. Our legal research tool is Lexus and they have built-in an AI, but I will tell you I have I've done some

[23:52] comparisons where I'll I'll redact all the sensitive information, but I'll run kind of a hypothetical or a scenario through and I'll ask a question cuz we like, "Hey, I heard about this new law in Oklahoma. Tell me about it." And I

[24:06] have to go research it. And so I'll plug it into you chat GPT. Perplexity is my favorite. And then there's two legal ones that we've tried. And Perplexity's answer is almost always the best. >> Mhm. Which is we're still on the free

[24:18] version. Like I'm not even paying for it. So it is wild and I think why Perplexity is so good, at least right now, is that it's pulling from articles. It's almost it's like a souped-up Google. So it is pulling from real

[24:30] making it up. It's getting a little better. And then even the the legal research version, it's trained to be so tight to the laws that like it isn't creative at all. So if you're not using the exact same words that it knows to

[24:45] uh so it's interesting how the different models are tuned and even like the affect the results that you're getting. So before we move on to this next big category, if we're talking about three big legal risk areas, we're talking

[24:59] about data privacy and client risk best practices. Like how do I apply this, you know, in this next month in my online business as I'm using these AI tools, Yeah, I think it really depends like do you have a big team or is it just you?

[25:15] If it's just you, your risk honestly is pretty low because you're in control. You know, you can test things. I would put some sort of if you're doing any client work, I would put some sort of general term in your client contract

[25:27] saying, you know, we use AI to improve customer experience. You don't have to say what tools you're using because if you're like us, you're testing new know, changing workflows all the time. So I would kind of just give your

[25:39] clients a heads up. Um if you do have any, again, I don't viewers with this, but if you do have any contracts with like big corporations or universities or like the federal government, they have a lot of rules. So

[25:52] your contract. They may not allow you to be using AI or they may have like really strict restrictions on it. >> Yeah, I was going to say that's a it's a brilliant tip to say for anybody that has clients of any kind to just add to

[26:06] your legal contracts that we use AI, get that language right. That's that's a sharp action item for today. Yeah, for sure. And then if you do have a team, even if it's just contractors, have a have a talk with them about what is okay

[26:20] encouraging them to experiment? Do you want to keep it a little bit tighter so that maybe you've got one person who's a designated AI tester and you're checking the privacy settings. It's very easy for there to be

[26:33] different tools without any oversight because it's moving so fast. So kind of think about how what kind of policy you want. Um we I'm like the cobbler's child who has no shoes. We're working on our AI policy here at the firm. I have a

[26:47] the first version that my team drafted it totally unworkable. It had all these training requirements. I was just like, "We're a small team. We are not going to >> So, just make sure it's reasonable for how you actually do business. And that's

[27:01] our third category. Just a few other things under contract and team risk. So, things under contract and team risk. So, for anybody that has a team, um what if a contractor is using AI without telling me? Yeah, if something goes wrong, it's

[27:13] probably going to be your problem, not just their problem. Um and so that is it's a pain, um it really does like the buck stops with the company, not the contractor. Um so, you want to make sure you at least

[27:28] doing. And if you want to set down some guardrails, um you you know, something intellectual property. Like we talked about at the beginning. They could be um shouldn't be. Feeding confidential information that you wouldn't want. Um

[27:44] even things like your client lists and things like that, you know, that you property, but it's still core to your business. You probably wouldn't want a business. You probably wouldn't want a competitor um you know, finding it or

[27:56] internal processes and things and they're just loading them up into they're just loading them up into ChatGPT to to produce copy. And now like all of your frameworks or your curriculum is just out there. How do you

[28:08] we have coaches listening to this? People that do different coaching, they do events, they do masterminds about this AI bots recording Zoom meetings or masterminds? >> Yeah, so we we use Fathom at our firm. I

[28:22] because I try to write really quickly, but if I'm in a conversation, it's very hard for me to do both well. Um so, I think they're very useful. I think this tools is you really have to look at the terms to see what they're doing with

[28:37] your data, how they're using it. There may be some things that you can toggle off and on. Uh we were just upload updating our Zoom settings. Now Zoom has a built-in AI companion. I'm still trying to figure out what exactly it

[28:49] does. But but it's it's there's a lot of different choices that you can make. But they're not always great about kind of have to dig a little bit. And

[29:01] I'm curious like so when when you're teaching proprietary processes or you know, we teach all kinds of stuff here on the Think Media podcast. But then we have levels of coaching. We have a group coaching program, one-on-one coaching.

[29:14] someone come I mean I guess they they can be there and take notes, but now their AI bot is joining the Zoom meeting as well. What's your thoughts? I do not First of all, just as a participant, you see all these like black screens, the AI

[29:29] note taker in the room. So from a participant level, it's not great. But you have no idea what they're going to do with that information. They could publish it, they could use it to you know, create their own GPT to use

[29:43] your framework to do something else. And then there's also confidentiality issues. So often on these group calls, people are being vulnerable, maybe haven't shared publicly. They can be really emotional spaces. And

[29:56] so the idea that you could have 15 note takers in a group call and any one of them could you know, accidentally upload or a team member could access it that wasn't supposed to. It's just a lot of risk. So we when

[30:10] contracts, we'll often put in there for group coaching programs like they are have to remind people at the top of a call if there is a set to join everyone please turn off your note takers or you can just kick them out.

[30:25] It definitely opens the door to problems for sure. So we want to hit this gold framework that you have in the middle and I have a couple fast questions, fast to make sure that listeners get handles. And so we have a

[30:41] checklist coming towards the end. Um but if you're listening to this right now and you are thinking I might not have this handled, Autumn actually teaches uh this down inside of what's called Fine Print Academy. Now, if you want to check

[30:57] in the YouTube description and in the podcast show notes because really your next mistake um could be costly or the thing you don't know could be very costly. But Autumn, what exactly is that program and who is

[31:10] it right for, you know, to save us uh time, energy, risk, and protect our Print Academy is really intended for folks maybe you haven't even started your business yet up to about half a million dollars annual revenue. And

[31:24] afford to have an attorney in their back pocket that they can call anytime or to hire an attorney to draft a contract. Um so, we help them set up a revenue protection system. So, every dollar that you're coming in the door is protected

[31:37] by a solid contract, by a solid refund policy. You can kind of breathe a sigh protected as you're starting to work with clients or sell a coaching program. protection system. After you get that, after you get your revenue all

[31:51] intellectual property, setting up a business entity, kind of all of that other stuff that goes into having a legally protected business. Uh but it's very easy. We have human support. Of course, we do use AI behind the scenes

[32:04] to make it flow smoothly. Um but you get contract templates and real human support. That's powerful. What does a the typical range of hiring attorney cost per hour? Uh I mean, I'm 450 $450 an hour.

[32:17] hired our firm it's going to be probably about $5,000. So, it's it's a significant investment. >> So, yeah, hiring an attorney 500 bucks an hour, getting one contract fully locked down $5,000.

[32:30] Uh Fine Print Academy is very affordable when you think about it being a resource keeping it updated and you're staying on the cutting edges, right? whole part of the market, all these business owners who really weren't being

[32:45] served. It was basically either you can go to LegalZoom, you can buy a template, Um so it's meant to really have a little bit more hand-holding. We at the law firm get inquiries every day that we're not a good fit to help and it just felt

[32:58] cold. So we designed it to be really affordable, really accessible. Brilliant. So if you uh want to check out Fine Print Academy, just go to that link in the show notes. Um and I want to talk about you

[33:12] mentioned this earlier, but this is your human in the middle rule. This is like a branded framework for you and that's use AI at the beginning and end, but human in the middle. Can you expand that as why this is a best practice? Yeah, so

[33:24] were talking about copyright protection. How can you make sure that you actually own the copyright in what you're building, which is use AI for brainstorming, use it for outlining, use it for, you know, give me 20 hooks, give

[33:38] me some ideas, um give me the topics that the most know, whatever social media platform so you can make sure you're staying on top of trends. But then you really need to use your own brain in the middle and

[33:51] bring your personal touch, your human, your ideas, um so that you're writing an outline. It's fine to use it so you're not staring at that blank page with the cursor blinking at you. And then at the end, again, to wrap up, to

[34:03] make sure, you know, a good proofread. Am I using language that's appropriate to? You know, am I using too much industry jargon, which I'm the worst at. Um you know, that kind of final check um or I like to use it to kind of soup

[34:17] um you know, all of the things that we do to refine our work before it's finished. And then also with the nice thing with video content is that the script is not the final thing. Like you're

[34:29] talking, you're it's a whole 'nother, you know, {quote} work when you create the video. Um so, you can add more human touch there as well. Ideally, I don't think we want to be reading a script verbatim. Um so, that

[34:41] you got the human at the beginning and or sorry, you've got AI at the beginning middle is where like all the magic happens. That's brilliant. All right, I questions. We've touched on a few, but I mean, you you could say yes or no. You

[34:54] might be tempted to expand, but can we use AI for voice-overs? Sure. Can we use AI for thumbnails? Yes. Can we use AI to write cease and desist

[35:08] Uh you can. It's not a good idea. Why is it not a good idea? Uh because the lawyer that gets it is going to know it was AI. Mhm. Can we I mean, is that what you also do in the academy? Like you at least could have

[35:22] stronger or you have good templates? Yeah, we've got great templates and we'll give things a once-over. Yeah. Uh Make sure people aren't saying crazy see a red flag. That's brilliant. Can you use AI to search trademarks? No.

[35:40] Uh so, AI again is generative. It's not a real search tool. Something like a Perplexity might be okay, but um it's not going to catch things that a real trademark search will catch like sound-alikes. Um it's not going to

[35:54] remember, AI is always based on past things. Um it's and it's it's not going to understand the nuances of trademark law, I think is the biggest one. Mhm. So, um and I think you can actually search trademarks at the

[36:09] United States Patent and Trademark Office. Correct. Um and so, but you would you're the there's a really big insight. AI is generative. I suppose there's always that discla- disclaimer where it's like, check

[36:21] things. AI might hallucinate. And so, the simple truth is it does. than just making things up, it misses things. things. Can I use AI to search contracts?

[36:35] best uses of AI. Or like to help you understand a contract that you >> So if you get a contract, put that in there, explain this to me, point out red >> Yeah, what might be disadvantageous to me. Mhm.

[36:48] And then can I use AI for hiring? No. Why? Uh AI has uh it was built by people who have uh who are generally, no offense Sean, who are generally white dudes. And

[37:01] so it has a lot of built-in bias. Mhm. Um and there are lawsuits pending right now with people saying that um you know AI tools that review resumes were, you know, biased against certain groups or certain protected classes. Wow.

[37:16] Yeah. Okay. >> It opens you up to potential problems. Got it. And And would your framework help though if it's like AI at the beginning, human in the middle, and AI at the end? Or you just want to be

[37:29] >> Yeah, I think you want to be pretty careful. Um I would say the the human in the middle is great for content creation. for legal compliance. Okay, so before we wrap, um

[37:45] is there like a checklist of things we should do this week that is kind of based on um next steps so we leave this episode with some action items? Yeah, number one, look at your contracts if you're

[37:58] selling anything, whether it's coaching, you're doing client work, basically on YouTube. Uh check your contracts see if it deals with AI, if not, it's a great time to add it. I would also, while you're in there, look and make

[38:11] you actually do business cuz I find a lot of people just set their contracts often and then you look at you're like, "Oh my gosh, we update our contract." And then number two, take a look at whatever are kind of

[38:24] day-to-day in your business, and check your privacy settings. They may have changed. Uh there may be some options that are there now. Um and just think about what kind of data are you putting

[38:37] in there. It may be like if you're doing mostly marketing, it's probably fine. of financials or anything that touches on client data, things that are more those settings. And maybe number three,

[38:50] those settings. And maybe number three, if you have a team, should you implement or review your AI policies and workflows? Yeah, at the very least I everyone and just ask what people are using. That's where we started. Mhm. Um

[39:05] everyone's like, "I don't like it. I'm not using anything." We're like, "Great, have a more creative team, you will probably find a lot of things you had no kind of go from there and figure out what kind of policy you need. So, AI

[39:20] isn't illegal, but using it blindly can be dangerous. And you've given us a lot of tips here. I want you to shout out your stuff, but I also uh if people can with you, um as well as remind people though that

[39:34] if you're listening to this, especially if you've been guessing on um and you don't want to wait until you get a scary email or something happening legally that could be unfortunate or costly, then Autumn's Fine Print Academy

[39:49] contracts, IP protection, compliance for digital businesses. So, there's a link to that in the show notes and the to follow you as well, we'll make sure that all of your stuff is down there.

[40:01] Where can they connect with you? Great. We are AWB Firm, which is my initials. website. That's our social media handle. That is our YouTube channel. Thank you, Autumn.