Are you legally compliant with OpenAI and the products you create?

Welcome to another episode on the FSBO Podcast. This is the Small Business Credit Minute, and I'm your host and business credit expert, S.E. Day. Over the past few episodes, we've been talking about artificial intelligence, and today we're diving into a topic I've mentioned on several occasions, generative AI, and more specifically, how it affects small business operations, productivity, and scaling. To help us get this right, we'll be joined after the break by attorney Michael McCready of the McCready Law Firm based in Chicago, Illinois. Stay tuned and we'll be right back. In today's competitive world, your business deserves more than just hard work. It deserves the financial freedom to grow. Building business credit is the key to unlocking opportunities. With strong business credit, you can expand faster, invest smarter, and reduce your personal risk. Start building your business credit today because your dreams deserve the chance to grow just like your business. Build your business credit. Secure your spot at the next Small Business Credit Survival Bootcamp.

Hey, good morning.

Good morning. Welcome back to the show. As I mentioned before the break, Attorney Michael McCready is joining the show to talk about generative AI and how it affects operations, productivity, and scaling for your business. Michael, welcome to the show.

Good morning. Thanks for having me. Looking forward to it.

Absolutely. It's always a pleasure to have a guest in here who's knowledgeable to talk about these things. So as we dive into it, Michael, what exactly is generative AI? We've been talking about this on a few episodes and building up to this where I wanted to lead off with— end it off with a legal eagle on this stuff and why it matters to small businesses. So what is generative AI?

Listen, the term gets thrown around a lot, but at its core, it's essentially a software that can create something. Okay, it can create text, it can create images, audio, it can create code, any number of things. But it's a tool that businesses can use to create something.

Right. Okay, so with that being said, Why does it matter now?

Unless you've been living under a rock, AI is everywhere. And you may not even realize where AI is, but AI is already embedded into a lot of the programs that you already use. It's in QuickBooks, it's in Microsoft, it's in Google, it's in so many places. And you, if you're not utilizing these built-in AI functions, you should. But where the real benefit is, is using AI in other aspects of your business.

Okay. Okay. So is AI now today more of an opportunity or more of a risk?

Oh, it's absolutely an opportunity. The people that refuse to embrace it for whatever reason are going to be left behind. I hear frequently that, oh, AI makes mistakes. Well, you know what? Humans make mistakes. The difference is that if AI makes mistakes, you can teach it and it will learn and it will not make that mistake again. You can't say the same thing about humans. How many humans make the same mistake again and again? So, you know, I've never found that to be a persuasive argument.

Got it. Got it. So from an operations standpoint for a business, you and I are both business owners, been at this for some time. Where are businesses using AI from an operational standpoint that you see today?

I think it's important not to get overwhelmed with all of the options out there. You know, start with the low-hanging fruit. You don't need to incorporate AI into every aspect of your small business. So start with the things that are easy. I mean, AI can It's great at drafting summaries.

Mm-hmm.

It's great at doing proposals. It's great at doing marketing, job descriptions. You know, all of these kind of things are really easy to do on a one-off basis. And then once you become comfortable with using AI in that situation, you know, then I recommend looking at, you know, each aspect of your business and decide, you know, how could AI be applied to this? Here's an example in my own law firm is we've got a series of communications that go out to a client over the course of the case, explaining what's going on and so forth and so on. I mean, I've been using those same forms for years. We ran them through AI with the prompt is, please reword these so they are understandable for somebody with an 8th grade education.

Ah.

All right. And it reworded all of our correspondence. It sounded less like lawyers, and it sounded like, you know, what people would understand. And the reason I said 8th grade, and this is a good tip, is that, you know, not everybody is college educated.

Right. Yeah.

You said, please reword these for someone with a PhD. It's going to give you a different answer. So we found with client communications is if you prompt it to generate it based on someone who's got an 8th grade education.

And basically what you're saying, Michael, is the literacy rate— I come from the public affairs arena, the public relations. The literacy rate in this country is you have people that are on an 8th to 9th grade reading level. So I understand when you say make it where an 8th grader can understand it.

Totally get it. Regardless of what their education actually is.

Yes. Yes.

There are a lot of people that read at the 8th grade level.

Absolutely. So, Michael, what's the danger of employees using AI without some type of guidance? So in your law firm, for example, what's the danger in that if they don't have some type of guidance in place?

Yes. Absolutely. AI is a tool, okay? And it is not the beginning, the middle, and the end. And so in our firm, it is AI can create, but a human has to verify, and then AI can polish, okay? There has to be that human in the middle. You absolutely cannot leave AI onto its own. And the second thing, and I know we'll dive into this more deeply, but, you know, there are things that AI is good at and there are things that AI should not be used for. And you do need to make sure that everyone in your business who has access to your data understands the difference.

And speaking of data, what about confidentiality and sensitive data?

So, you know, we're using the term AI and it's very, very broad.

Right.

You know, generally speaking, we're talking about LLMs, large language models, such as ChatGPT or Claude Anthropic or Gemini, Grok. So there's a handful of these. Each of them does have the ability to control the privacy settings. And it's absolutely essential that Every business owner takes the affirmative step with everyone in their office to make sure these privacy settings are set properly, because by default, you know, AI and the and ChatGPT they learn from the data that you give it. That's why you know AI is is generative. It keeps going and keeps learning. You don't want that. You you don't want it being trained on your data because your data is. It could be sensitive, it could be financial, it could be any number of things. So each of these programs has got a privacy setting that allows you to turn it off and not share your data with the AI. And you need to make sure that everybody is aware of that and everybody does that.

So Michael, what you're saying is that leads right into a policy, for instance, an AI policy for your company. What type of things should be in that? In that policy, in that AI guidance?

You know, a lot depends on what your business is. You know, there are certain businesses that are highly regulated, and I'm not telling you, you know, what should or shouldn't be, but that may be a concern. The law is regulated, so there are things that we don't use AI for. You absolutely need to make sure that there's human in the loop, that somebody verifies it, In fact, we have, you know, we basically have an audit trail that people have to sign off who reviewed this before it goes out the door. In that way, we have accountability. And then, you know, there should be someone who is monitoring compliance. I'd love to say that every employee follows every rule, but we all know that. And so we all know that not to be the case. So, you know, I encourage, you know, business owners to create some kind of compliance monitoring. And then finally, as the owner, as the leader, as the visionary, you know, you should lead by example and make sure that you encourage people to use AI. You don't want people to be nervous of it or scared or be too intimidated. This is a tool that everyone should use, but they do need to be trained and understand what the, you know, the guide rails are going to be.

Michael, on our last episode, I had a young lady on and she's an AI integration specialist. And she was touching a lot on what you were saying there, whereas you should have your employees trained. She said that she, as a trainer, one of the things that she does is she encourages an AI training day for the staff, for the leaders, for the owners of these businesses. That way everybody can feel a little bit more comfortable with AI instead of just plugging stuff in and playing it without reviewing it.

Yeah, I'd agree with that. Something else that I think that all small business owners should learn, this skill, is it's called change management.

Okay.

And you can't just expect your employees to all of a sudden embrace some, you know, very big changes. Oh, look, we've got AI. Okay, everybody use it. No, you need to, you need to learn how to gradually roll these things out. There is a science to change management.

Yes.

And another criticism that a lot of companies are saying is, we have all this AI and it's not really, you know, helping us at all. Well, it's probably because you didn't roll it out properly and people aren't using it properly.

Right, exactly. So what's the risk in letting AI replace quality controls, if you will?

I don't think that it can replace. I think it supplements. I think it augments. Okay, AI shouldn't replace anything. Now, AI can do quality control.

Right.

If it's trained properly. And once again, your listeners are in a whole bunch of different types of businesses. So it's hard to say universally. So you can use it for quality control, but you have to leave that human in the loop. You have to have the final approval done by humans.

And, you know, to add on to that, you know, AI is confident, right? When it spits something out, it doesn't think it's wrong.

That's right. That's right. So here's a, you know, and stop me if another guest has already talked about this, but this is how hallucinations happen. Okay. Is AI does not think. Okay. AI predicts.

Yes.

And based on such a huge volume of data, it can make very, very accurate predictions and that, and you get your answers. But sometimes when you don't When AI doesn't have enough data to create those connections, the predictions, you know, they can't do it. And AI is trained to not admit that they don't know.

Right.

And the analogy is when you take a multiple-choice exam, right, and you don't know the answer, right, you can say, okay, A is definitely not it, right? It could be B, it could be D. Yes, because if you leave it blank, you're going to get it wrong.

Right, exactly.

That's what AI is doing. Okay. It's making an educated guess based on the information that it has. And if that information is not sufficient to come up with something that's accurate, that's when it hallucinates.

You know, and you said something, and this has kind of been the running theme of prompting. You got to prompt it correctly, right? And the other side is you got to review it. You have to review it. But Michael, this is something that has been on my mind for a while. When AI creates something, who owns that content?

Generally speaking, the business does. Okay. Well, let me backtrack. Okay. The person who creates the AI-generated content arguably could be considered the owner of that content.

Right.

Now, usually if you're an employee, then that goes to the employer. Not always. Sometimes in certain businesses, you need a contract with your employee that says anything that you create belongs to the business.

Right.

Think of artists, right? You are an artist and you work for me. You know, I own your creative output because by default the law says that the artist owns it. So AI, you know, I think that— so that's a caveat. But generally speaking, AI is owned by the business unless the AI which you are using has some kind of contractual provision. Now, you know, the big ones, ChatGPT and others, they don't. But if you're using a custom, you know, a small vendor, an AI vendor, that is something that you need to be concerned about. There are other concerns about, you know, vendor contracts, if you want to talk things that you should investigate before you employ AI and a program in your business.

And that's very, that is very crucial because now people are just, especially through the generative AI model, they are creating things. Um, some of it is done on free sites, some of it done on paid sites. And when that content comes out, because a lot of the AI, a lot of the content that you input into the AI, they're training their models on.

Right.

So they still have access to anything that you've created there. So it's gonna get tricky from a, you know, from a legal standpoint.

So listen, something I've told my kids is that nothing in life is free, kids. Okay? If you're not paying for it, you are the product.

Yeah, that's it.

So, so, uh, so please don't be fooled by free anything when it comes to AI. And data is important. And you want, like, getting back to what I said earlier, is you want to make sure that you maintain complete control over your data and your data that you're using in AI is not being used to help improve the program, which only benefits the program owner. It doesn't necessarily benefit you.

Exactly. Now, Michael, you've touched on this. How is it that you can Train your staff to not be over-reliant on AI. I mean, what do you do? Because that's a— that can be a problem.

Yeah, I'll be frank with you. I don't have an answer. It is so— I mean, even me personally, you know, I rely on it so much, almost to the extent of over-reliance. There's There's lots of literature out there that AI can do so much now that it's limiting people's own ability to think creatively and to process information. And I think that's going to continue to be a problem. But as long as you have a policy about what can and can't be used— listen, when I see something written to me that's clearly written, by AI.

Yeah.

You know, I'm not insulted by that. That's a tool. I want people to do that. I think that some people are being, oh, they just ran it through ChatGPT and this. Well, if they did or they didn't, it doesn't matter. This is the right response.

Right.

So, you know, and we've got— I've got employees that English is not their first language. And AI is great. You know, they can write what their response is to a client. Run it through AI and it will clean it up really nicely. That's a great use for it. And if they do that with every single correspondence, I think that's great. 100% support that.

So in supporting that, now we're going to shift gears a little bit into the scaling side. So how is AI becoming a part of the growth infrastructure? Businesses like yours. I mean, I'm hearing you talk. You are a barred attorney. that practices law, but you're also a business owner. And the way you're speaking, I love it. You're speaking from a business owner standpoint with the legal cap on over here, you know, from the legal aspects and understanding how this is all tying into things. So how is that becoming a part of the growth infrastructure for businesses?

You know, listen, there, there are core principles, core business principles that apply regardless of what your business is. That's why you have people that go for MBAs.

Yeah.

Right? I mean, there are certain best practices, and AI is rapidly becoming a business tool that doesn't matter what business you're in, you have to master it, and you have to master its application to your business and your product or your service. So yes, but, you know, I am in the service business. My services happen to be legal services.

Right.

But first and foremost, we are a business and we are in the customer service business and our customers are our clients and our service is legal services. And so from that perspective, you know, these principles, business principles, AI principles can be applied universally.

Got it. So now, Michael, in a small business, a small business owner is looking at AI. They've been listening to the show. Now they're ready to dive into it. What are some of the, what are some of the contract clauses that matter when signing up for AI tools?

One is who owns the data, which we touched on, right? They're going to be used to train the other system. You know, other important things are confidentiality, liability. Okay, I don't think you can do much about this. Almost every contract that you sign, they're going to say we're not responsible if anything goes wrong.

So that was my next question. Who is liable if AI goes wrong in giving bad advice?

Yeah, and I mean, even if you look at ChatGPT, absolutely every single result, there's a little thing at the bottom is—

I forget. A little disclaimer.

A little disclaimer that says that ChatGPT could be wrong. This should be verified.

Exactly.

You know, so from that perspective, um, you just, uh, you know, just— it's, it's, it's go slowly, right? Yeah, go slowly but go.

Okay, right.

You need, you need to be, you need to be on this and you need to be on this now.

So we've touched on a few things about, you know, compliance and training. How can a small business stay compliant with evolving AI regulations? Because you and I both know the laws and the policies have not quite caught up with AI yet. So how can they stay compliant?

So my first advice is use AI to figure out how to stay compliant.

Okay?

Use AI to come up with an AI use policy.

Wow. Wow.

So, and that goes back to the proper query, the proper prompt. So, I am a small business owner in the business of X. I have this many employees. You know, and be specific about what your business is. I envision my business using AI to do this, this, this, and this. Please draft me an AI use policy that I can introduce to my employees. Next, please give me a 10-step route— way to introduce it.

Right.

You know, what should I do first? What should I say? Who should I give it to? How should I roll this out? So use AI to implement your AI policies.

That is perfect advice. It's AI. We already brag on it of how great it is. Use it to help you out.

Yes, yes. You know, please, you know, I'm a small business owner. My employees know nothing about AI. Please give me a 3-week dedicated introduction on what I should do step by step to get my employees up to speed so I can roll out AI usage in my business. It's going to give you that.

Now, Michael, taking just a little deviation from a legal standpoint, one of the things that I believe in, I truly believe in, is that every small business should have a CPA. They should have a financial advisor, and they should have a legal expert on their team. That should be a part of their advisory team, right?

Mm-hmm.

So in them creating these documents, telling AI to create me an AI use policy, you should have a legal person on your team or on retainer that you can call that can review policies and contracts to make sure you're staying on point. So that's my belief.

What do you think? AI creates, but humans verify. Exactly. So AI can create this policy for you, but, you know, you or someone else, you know, really needs to make sure that it's what you need. Now, when I said an AI use policy, This is not heavy-duty law stuff, right?

Right.

You know, if you're talking about how do I comply with the TCPA, you know, Act or something like that, I would not rely on legal advice from ChatGPT.

Got it.

But it is really good, especially for creative things. If you've got SOPs, standard operating procedures, run those through AI and tell them to improve it. You know, what, what am I missing? What steps need to be clarified? You know what your SOP should be and what the result will be, and, and, and AI will certainly be able to clear or clean those kind of things up. So, you know, once again, AI can create, but humans verify.

So, Mike, what is one task that AI should not do without an attorney's review?

Let's see. Once again, everybody's business is different, so I don't know about compliance, but universally there are potential employment issues.

Yeah.

Okay. How you deal with employees, and I don't care if you've got 1 employee or 100 employees, you know, relying on AI to craft an employment strategy or You know, I've got an employee that's doing X, Y, and Z, and they just told me they're pregnant. You know, that's a potential real problem.

Right.

And I would not rely on AI to tell you how to address that. Now, I would rely on AI. I've got an employee who's chronically late, and I think that I want to fire them.

Yeah.

Please give me, you know, a detailed agenda on what I need to do in order to avoid a potential claim for unemployment or discrimination or whatever. So, you know, and it's going to tell you, you know, first is a verbal warning, second is a written warning, third is, you know, and that would be okay. But employment issues, I say, are something that Do you have to be careful about with AI?

Okay. So what's the worst AI mistake a small business can make?

Fortunately, I didn't make this mistake, but it was close. So we've been heavily involved. Well, we started with AI 2 years ago. So we're really involved now. But 2 years ago, I was so excited. I'm like, I've had my business for 25 years. I'm gonna dump everything into AI.

Right, right.

I've ever done. And, um, and then, uh, someone brought it to my attention. Well, if you do that, somebody could do a query. What does Michael really think of my work ethic?

Yeah.

And it's gonna come out. Wow.

Yeah.

And so, uh, so that could have been an absolute disaster. As you become more, uh, the more you utilize AI, Another thing that you're going to want to take into consideration is data, is protective, putting different data in silos. So, for example, not everybody in the firm should have access to the HR. Okay?

True, true.

Not everybody should have access to the financial. So, you know, creating data segregation and what is available to different people is, As you start adding AI tools and incorporating more, that becomes much more important.

Gotcha. You know, it's almost like a bad— this is like a bad analogy. Like you say, you, you see the disclaimers of drink responsibly. Yeah. AI responsibly.

Yeah.

That's what you should be doing, right? Yep. Michael. Yeah.

But, but, but I don't want to— I don't want to discourage people. Right? I don't want you to feel like there's all these pitfalls and these mines and everything can go wrong and I don't know where to start. And, you know, just start with a few real simple things. Teach yourself how to prompt, you know, how to make a query, because a query in AI is different than a query in Google. You know, we've become so accustomed just to typing in natural language in Google And what does Google give us? Google gives us a website that is responsive to our question.

Okay.

And I'm not talking about Google AI, but Google traditionally directs you to a website that is responsive to your question. And they've made a business model on an algorithm that's pretty good. AI is different. AI gives you the answer.

Yes.

It synthesizes and processes all this information and gives you the answer. So it stands to reason that the way you ask a question is going to differ from Google and AI. So, you know, go on to— go on to YouTube. There are countless videos on how to create prompts and teach yourself that.

Absolutely. Some of the best advice you're going to get, people. Michael, this has been excellent insight. Listen, before we wrap up, I understand you're based in Chicago and other places. How can people reach out to you? How can they contact you?

Sure, you know, my website is McCreadyLaw.com. You know, although I am a lawyer, most people, you know, understand that lawyers specialize in different things, and I specialize in personal injury cases. So, you know, I'd appreciate— you know, I'd love to help somebody on a business, but it's just not my expertise. But I would also suggest is that I'm very active on LinkedIn. And I talk a lot about my, you know, my philosophy and what we're doing on AI and changes and so forth and so on. And whether it's following me or some other thought leaders on AI, LinkedIn is a really good resource for that.

Well, Michael, I can tell you I will be connecting with you and we'll be talking more on this because this is, you know, we're 2, 3 years in now from OpenAI launching. This is just the beginning. This is just the beginning, and we're nowhere near where we're going to go with AI.

How about this? This is my disclaimer. Whatever I told you today may not be true in a week.

Totally agree. Totally agree. Hey, Michael, again, thank you for being on and just helping us understand some of this from not just a small business owner standpoint, but from a legal mind standpoint too. And you really put it out in everyday language, right? That, that every small business owner can understand. You didn't throw all the legalese out there. You put it in everyday terms, but you're speaking from someone who's knowledgeable about it, but also someone who's experiencing it in real time.

Yes. Thank you. Thank you. I appreciate that.

Absolutely. And audience, listen, if you're serious about making your business bankable, creditworthy— this is a Small Business Credit Minute— and AI-ready. Stay tuned to us and we will be bringing you more episodes that can benefit you in the growth of your business. And always remember, your dreams deserve the right to grow, just like your business. Until next time, thanks. Thank you.