707 AI in IT Contracts: What MSPs Need to Know

[Uncle Marv]
Hello friends, Uncle Marv here with another episode of the IIT Business Podcast, powered by NetAlly, the show for IT professionals, managed service providers, and others, where we help you run your business better, smarter, and faster. This is our Wednesday live show. Anybody that is watching here, we invite you to join in, in our chat section and ask questions, comments, be nice, be nice, is all I have to say about that.

Joining me today, Rob Scott of Monjur. Rob, how are you?

[Rob Scott]
I'm good, Marv. Thank you for having me. How are you doing today?

[Uncle Marv]
I am doing good and right out of the box, look at that, Rob is the man, Matt, in the chat.

[Rob Scott]
Thank you. He gets paid to say that.

[Uncle Marv]
So this is one of those conversations that I'm sure is being had all over the place by companies all over, and it's not just managed service providers, it's probably every industry. What can AI do to make us better when it comes to not just helping us draft an email or put together a presentation, but can AI actually help us write contracts? I'm going to throw a couple of absurd statements out there.

You and I, when we first thought about this, the question came up, can AI keep us out of jail? And that's kind of how we started with this. But then we dialed it back and thought about, well, is AI at the point where we can use it to write our managed service agreements, our statements of work, and all of that stuff?

So let me just throw it out there. What are your first thoughts when this topic comes up?

[Rob Scott]
I think AI is transformational technology. You know, I'm old enough, Marv, to have seen the typewriter give way to the word processor. I saw the first fax machines.

I saw the first texting machines. I had an incredible tool called a BlackBerry, which you could hold in your hand, get emails. But nothing that I've seen has the transformational capability in terms of technology as the large language models that we've seen emerge in the last couple of years.

And because of that, I think that the AIs now are capable of passing the bar in every jurisdiction and getting a license in none. So the state of affairs is that we're stuck with licensed lawyers. And the question is, how will the legal industry and its clients learn how to use AI to enhance the attorney-client relationship and deliver legal services more cost effectively?

I don't see AI as an option between lawyers and tools. In fact, the AI tells you, don't rely on this. Go see a lawyer, just like every vendor who gives you templates says, go to a lawyer.

So for me, AI is not an alternative to lawyers, and it's not in the near term going to be taking the job of lawyers. But I will say this is true in law as in other industries. Lawyers who use AI will be taking the jobs of those who don't very quickly.

[Uncle Marv]
Interesting. As you mentioned that. So a lot of my clients are law firms, and we've already been asked the question about how to properly use AI, specifically Copilot, because my clients are 365 driven, and talked about can we use Copilot to help us do things?

They are doing them in some areas of summarizing depositions and things like that. They're using them to... They may get PDF documents, medical summaries and stuff, and they may throw it into AI to generate summaries there, but it's a little different when you're talking about actually generating documents, whether they're from scratch or not.

You had mentioned templates earlier, and I was going to ask, this is just an evolution of that, right? In a sense that for companies that don't want to spend the money to have an attorney draft something, we've had templates in our industry for many years, whether they come from a vendor, from a group, from a coach, from another provider that is sharing their copy with you. It's really...

Is it much different than that?

[Rob Scott]
I think in some ways it's more risky to use an AI generated document than an industry provided MSA. That's saying a lot, since everyone who knows me knows that I've been crusading against the, here's a template, wink, wink, take it to your lawyer, and then people don't, and wind up with all kinds of terrible documents. It impacts me and my work, because I've got to go to court on all these bad contracts.

Makes my job more difficult in a lot of ways. It's sad, because nobody is protecting the MSP. There's no relationship being created.

There's no recourse if something goes wrong. My biggest fear is that these AI tools are at a point where they're not robust enough to develop the type of legal protections that are necessary, specifically in AI and IT services related licensing agreements and service contracts. I most recently demonstrated this when we hired a new account executive.

His name is Aaron, and Aaron is a former football player in Division 1. He is a very smart guy with background in software sales. He came from an enterprise software company before he came to Monjur, and he asked me, he said, Rob, you know, I was wondering, why do people use Monjur instead of just going to the internet and doing an AI query for, you know, build me an MSA?

And I wanted to answer his question, but I thought the better training opportunity was to demonstrate to him what I'm talking about. So we open chat GTP, and I put a prompt in that said, I need a very well-written, very detailed MSA appropriate for the IT managed services industry. And then we looked at the output, and we compared it, for example, to Monjur's MSA, which those of you who don't know what Monjur does, we're a legal service for MSPs, and for that reason, we developed over two decades a set of legal agreements that we use to assist our clients with customer contracting.

We made a comparison between that MSA that we use with our clients and the one that came out of chat GPT, and there was a wild gap, huge gap, between the legal protections that the Monjur platform, the human created, the experienced law firm created document, and the one that came out of chat GPT. Not even a close question as whether or not the chat GPT generated one would offer a similar level of protection.

[Uncle Marv]
Well, can I ask what some of those gaps are? Are we talking where they left out certain provisions when it comes to services, or was it stuff that had to do with liabilities? I mean, what types of gaps are we talking?

[Rob Scott]
Yeah, so in the MSA, we wouldn't be talking about services at all. You know, typical structure more for an MSA is to deal with the general business terms and leave the order level service either attachments or service level agreements to hang off of. So we're talking about specific legal protections.

Things that are designed to protect the MSP that are in the nature of general terms and conditions, risk balancing issues, those types of things, as opposed to service descriptions.

[Uncle Marv]
Which would, those would be more in your statement of work and stuff, but normally, when you talk about descriptions and liabilities, I mean, there's a lot of things, you know, what are the responsibilities of the client? What are the responsibilities of the IT service provider? Are we talking about gaps in those things?

[Rob Scott]
A hundred percent. Okay. Gaps and exclusions, gaps in customer obligations.

Those would be two big categories where there were gaps.

[Uncle Marv]
So let me ask this question, because in the same vein as a template, whether you present a template or you present something drafted by an attorney, if you put it in front of a client and they sign it, in a sense, you have an agreement. So the question is, are those types of agreements, whether they're template, AI generated, are those things legally valid and, you know, enforceable by a court?

[Rob Scott]
Yeah, I would say that the court would likely not give much consideration to whether or not the contract itself was written by an attorney or written by AI. I don't think the method of generating the agreement itself would be an issue of enforceability or interpretation in court.

[Uncle Marv]
So it would be, well, you know, if you left out these provisions, doesn't matter. If an attorney leaves out the provisions, at least you can point to the attorney and said, you know, this was done incorrectly. But if somebody generates it by AI, I mean, who do you point the finger at?

[Rob Scott]
I mean, this is the key issue. I mean, this is the point that I was making earlier. It's like the AI can pass the bar in every jurisdiction and get a license in none, can get no professional malpractice insurance in any jurisdiction.

So in the end, you know, lawyers are going to stand behind the work, be responsible to the client for any negligence, will likely have professional liability insurance to protect the client from any such negligence. And otherwise, they're standing behind professionally what's in the documents. And I just don't know how.

And listen, I use AI every single day. When ChatGPT4 launched, I wrote a book with it within the first two weeks. I registered a copyright on an article entitled, How to Get a Copyright Registration for a Work Containing AI Generated Content.

And then I got a copyright on that work that I wrote with ChatGPT. So I'm not against AI. And I'm not suggesting that I'm not one of these lawyers that's like trying to hold on to the past and explain why AI is evil.

I'm a proponent of AI. I just think that AI is a tool for experienced lawyers to use to help deliver more cost effective, more responsive legal services. It's not an alternative to using lawyers.

[Uncle Marv]
All right. So let me ask this question. And this, I guess, goes to any agreement, whether somebody generates their own using AI.

They get a customer to sign it. They haven't, you know, consulted with their legal authority. A dispute comes up.

They come to you and say, hey, I've got this issue. Here's my contract. Can you help me enforce it?

Or can you defend me in court? I've actually never asked this question of any of the attorneys. Would you actually take that case after looking at the agreement?

Or would you say, yeah, this is Swiss cheese. This isn't going to do anything.

[Rob Scott]
Look, I think that even if there's a bad agreement, and I'll be honest with you, we've got cases right now all over the country where I'm not happy with the agreements. I would, you know, if we didn't write them, then I'm going to find fault with, you know, anything that doesn't give us the same advantage in court as what we would have provided. However, even in those cases where you have a bad agreement, there's usually a lot of issues to defend.

First, with respect to liability. Even if there's a bad agreement, the plaintiff still has to prove a breach of the agreement, which means that the MSP did something that it was contractually obligated to do or failed to do something. And therefore, there's a breach.

Even on a bad contract, you need a good lawyer to defend that case. Similarly, even in cases where it's clear, for example, that the MSP or the service provider was negligent, there's usually a big dispute for your defense attorney to argue about whether the damages that are being claimed in the case are appropriate. I'm working on a case right now where, you know, we represent the MSP in an arbitration in New York City, and just a wild difference in how the parties see what's recoverable under the terms of the contract.

So even if you have a poorly written contract, you still need a defense attorney. And I would say the worse your contract is, the better, more experienced the defense attorney you're going to need. Our agreements are designed to try to make it to where you'll have less cases because they're well-written.

And number two, that if there's ever a need to defend the contracts in court, that your insurance company will pick up the tab, not some unplanned, unbudgeted, huge expense for you with legal fees. The way Manger works is to work in synergy with insurance so that if there is such a claim, that your insurance company will pay the fees for experienced lawyers to defend you.

[Uncle Marv]
All right. So let me ask some other questions related to AI, because one of the things that keeps popping up is this intellectual property rights, like who owns the data of AI-generated content. Does that actually come into play if we're using AI to help us with our contracts?

[Rob Scott]
It's less important if you're using it to help with your contracts other than if you're involved in offering managed AI services, there's a whole set of new concepts that you need to cover that aren't covered in the traditional managed services agreement. It's for that reason that we launched our service attachment for managed AI in March of this year. You need a service attachment that covers the concepts of AI, not because you're using tools to generate contracts within your business or not because your stack of vendors has AI solutions incorporated.

You need a managed AI service attachment when you begin helping clients implement AI solutions within their organization, such as Microsoft Copilot, such as conversational voice solutions from companies like HATS AI. That's when you need to have a separate agreement. In that agreement, there's a number of intellectual property concepts.

You mentioned a couple. But when I first started studying AI, my first mission was to understand the intellectual property differences between a software transaction with AI technology in it versus a traditional SaaS agreement without AI. And we developed a number of what we called IP artifacts.

These are concepts that are unique to AI. And they include, as you mentioned, AI outputs, AI inputs, training, the model, the learnings. So all of these concepts are new to AI and create intellectual property challenges because under traditional software licensing agreements, all of the IP to be considered under the contract was already developed at the time of the license.

What's unique in AI is IP gets generated after the license agreement is entered into. And so it becomes very important to identify these items. And as the market is evolving more, the prevailing model, the prevailing wisdom is this.

The provider or the vendor owns the model. The input data and the output data belongs to the client. And the learnings should be discussed in a way that whatever learnings the vendor is getting aren't being used in a competitive way against the customer.

For example, if I've got an AI solution that is gathering data about warehouse operations and one of my clients is Amazon Warehouse, Amazon Warehouse isn't going to want me to be able to take all of their data about how they move around and what they do from a shipping and logistics perspective with other people in that space. They would consider that a competitive secret. This is an example why AI and IP at every level is such a challenging concept because it's so difficult to understand all of the uses of the data.

And that's why you see companies like Salesforce having as central to their AI marketing is that we're not going to steal your data. We're not, you know, you've got the cowboy Western scene, you know, we're not going to steal your data. And so data protection, data privacy, intellectual property is being turned on its head by AI.

And that's why you need specific agreements to address not only the IP issues, but for the same reasons you talked about earlier, Marv, a good service attachment for managed AI is going to have the right exclusions, the right client obligations. And the right risk balancing so that service providers can help clients, but for example, not be responsible if, for example, there's a regulation that comes down the pike that makes the investment in that project no longer viable because there could be regulatory changes. Or the concept of both parties being committed to responsible AI.

You know, I'm not a conspiracy theorist. I'm not that concerned that, you know, the AI tools are going to be telling me what to do. However, I am a firm believer that we all have a responsibility, including lawyers and developers and end users and vendors, to be committed to responsible AI.

And that commitment should be extended to a contractual commitment on both sides where both parties are so committed. And you'll see that in our managed AI service attachment.

[Uncle Marv]
All right. So we talked about these protections that have to be, you know, considered on both sides. Does AI-generated content fall under the same data protection guidelines that we need to do with customer data in other areas?

For instance, you know, with my law firms, if they get anything from opposing counsel, only certain people can have access to those cases. They have to do, you know, those checks, you know, conflict checks and stuff like that. Do we need to consider AI in that same vein with separation of who has access to the data that's generated?

[Rob Scott]
I doubt that you would see an attorney's eyes only type of requirement from an AI tool.

[Uncle Marv]
Okay.

[Rob Scott]
However, there could be copies of chat sessions or logs of chat sessions that become part of a case in the same way that an email or a text could become part of a case. And I can envision a situation where there were chat sessions that were requested in discovery that a party had to produce. And as a result of that, because it had certain information that it would claim is attorney's eyes only, which means the attorneys can see it, but the clients can't.

I can see it happening there. But only in that context of, you know, someone has some copies of some chat sessions or logs of chat sessions that become the subject of a dispute. However, the same privacy, security, and compliance concerns regarding data privacy and security from a regulatory perspective exist equally in AI applications as they do in SAS.

And for example, you can't put healthcare information in an open AI session without violating HIPAA. Now, law firms are building tools and Microsoft is building versions of Copilot and other AI tools that will allow law firms and others to have confidential repositories where the outputs and so forth for what they're using with Copilot are not being made publicly available and therefore would not violate attorney-client privilege or other issues. But privacy, security, confidentiality are still really big issues when it comes to AI, particularly in the context of law firms who have an obligation to all clients to safeguard confidential information, which is a more sacred rule in terms of attorney ethics than any data privacy rule that's been written.

And so law firms have a huge hill to climb in safely adopting AI tools because of that attorney-client privilege. The obligation to safeguard data and the fact that it's not abundantly clear from everybody who's using these tools exactly where the data is going and how it's being used. And so it's incumbent upon all law firms to have programs where they train their employees on how to use AI.

If they're going to make it available at all, then they should train on how to use that and how to make it to where it's being done appropriately. And I think that once the law firms have those platforms, you'll start to see them. They're available.

Those platforms are out there for lawyers. You're going to start seeing lawyers using AI more and more, back to my point from earlier, Mark. AI is not taking the job of any lawyers anytime soon.

But lawyers who don't use AI are going to be losing their jobs to lawyers who do quickly.

[Uncle Marv]
Right. So I know that one of the things that comes into play is these ethical and professional concerns. And the phrase that comes to mind is duty of care, where using AI-generated stuff without proper review may not meet the professional standards.

[Rob Scott]
I don't think there's any question about that.

[Uncle Marv]
But I mean, but that's going to apply more to attorneys than to us, because we're not lawyers. We would not, we can't hold ourselves up to a professional standard of an attorney. But this idea, there was a thing I read that talked about the unauthorized practice of law, where an AI-generated contract could fall under that for those very same reasons, right?

[Rob Scott]
I don't think that's likely to happen.

[Uncle Marv]
No? Okay.

[Rob Scott]
I think that if you think about the way the U.S. government is thinking about AI, is AI is not a person. For that reason, for example, the courts have said that AI can't be an artist on a copyright registration. So in my mind, an AI can't be a lawyer.

It can't be sued for professional responsibility. And it can't be sued for the unauthorized practice of law.

[Uncle Marv]
No, but AI couldn't. But me or giving this to an attorney and telling them to go into court with an AI-generated document, wouldn't the judge look at whoever's holding the document and say, I don't care that it's AI written. You're the one holding it.

You're the one that presented it. You're the one that's going to be held responsible, right?

[Rob Scott]
I think there's some leaps here, Marv, that don't, that add up. So let's take your hypothetical. We've got, you developed an AI contract.

You had the client sign it. You didn't run it by me first. Then you get sideways with the client and we have to go to court.

You're alleging the client owes you money, including an early termination fee. They're claiming you breached the agreement and you were negligent, and your negligence caused them to have a ransomware attack and they lost hundreds of thousands of dollars. So now we have a case.

So is this an arbitration case or a court case would be whatever your AI contract says. So let's assume it's a court case. I go file a case for you.

I'm required in the breach of contract case to attach the agreement. If the judge finds the agreement, doesn't, an interpretation of the agreement doesn't fit your breach of contract claim, that's not a reflection on me in any way. That's, you're going to lose your case because we didn't prove the breach of contract.

There's no way that a court is going to hold me responsible for an agreement that I didn't write. And the only thing that court is going to do is maybe rule against my client if I can't prove the case, which is the important point here, Mark. If it's below the standard of care for a lawyer to use an AI tool without proper supervision, what does that say about the propriety of a non-lawyer using AI without legal review?

I think the courts are going to say, you had an opportunity to hire a lawyer. The AI told you in the session that you should consult with a lawyer and you didn't. And if anything goes wrong, it's going to be on you.

It's not going to be the lawyer that tries to help you with your case afterward. It's not going to be the AI tool that said, this is just a starting point, go talk to your lawyer. And so therefore the only person that's going to get in trouble for using an AI tool in that fact pattern is you.

So that's why I was saying, the way you were saying it didn't quite add up.

[Uncle Marv]
But trust me, I'm kind of making this up because I have an attorney that does my stuff. But I'm just thinking of all the people that may say, I'll just throw it in the chat GPT or whatever AI tool. And it looks close enough, but you had mentioned earlier, yeah, just because it looks close doesn't mean it is.

[Rob Scott]
And yeah, and if you don't know what to evaluate it against, then are you really looking at it from a lens of a subject matter expert? That's my biggest concern. It's like, yeah, it types fast and it doesn't make any grammatical errors.

And when you read it, it sounds plausible. But when you're talking about complicated contracting between IT service providers and their customers involving regulated data and a lot of risk when it comes to cybersecurity, we're at a point now where there are legal solutions in the marketplace that are affordable even to the smallest MSPs and service providers.

[Uncle Marv]
All right, so we've spent a lot of time here talking about the risk and pitfalls of trying to do it yourself. So I want to take a quick break and then come back and talked about some of the best practices that we can do to get it right. But one thing I want to do is talk about TechCon Unplugged 2024.

I want to do it for two reasons. One, the conference is coming up in less than a month and I want to encourage people to go. It is September 12th through the 14th and it is in Hanover, Maryland, which is basically right on the other side of the city line of Washington, DC.

There's going to be over 25 educational sessions, two networking events and after hours entertainment, an all-inclusive expo with meals, refreshments and happy hours. I am going to be your MC. Not that that should be the reason for you to go, but just letting you know, I will be there.

And we've got some great speakers lined up, Marnie Stockman, Wes Spencer, Matt Rodella, Rob Ray, Amy Luby. And of course, my guest tonight, Rob Scott, is going to be there. You guys are a, what is it?

The Platinum Sponsor?

[Rob Scott]
Proud sponsors of TechCon. Yeah, we're excited to be there. Great to be mentioned with those legends.

You know, I know many of those speakers and know what they've contributed to our community. And it's an honor to be at an event and speaker with such a great lineup. I'm really excited to attend.

Monjur is really happy to be a Platinum Sponsor of the event. And we're looking forward to getting out and reconnecting and seeing everybody there.

[Uncle Marv]
All right, techconunplugged.com is the website. There are still spots available. Last week, when I did the show on that, I gave a discount code, which has been corrected because I did have some people try to use it.

If you are not yet registered and you're close by and want to get there, enter a discount code MSP75. That'll get you 75 bucks off the current ticket price. And if for some reason you are like 75 bucks isn't enough, I need more, reach out to me and I can probably get somebody there and get that money thing taken care of.

So if money is the issue, money is not the issue. So get signed up there, TechCon Unplugged and support my buddies, Parker LeBron and Rich Smith as they host that event. As I mentioned at the top of the show, we are powered by NetAlly, the network testing and analysis experts.

Check them out. They actually have a new tool I'm trying to get my hands on. They have a CyberScope Air.

Don't know much about it. I think it's just probably a Wi-Fi version of the CyberScope, but I'm reaching out to them. And as soon as I find out about that, I will let you know.

Our show is also sponsored by Super Ops. They are my mug sponsor here for the evening. If you're watching the live show, you see me drinking out of that from time to time.

They are also sponsors of the Florida Man segment that will be coming up. Elevate your MSP with the power of Super Ops. And our third partner for this evening, secure remote access to your windows and desktops.

Get it done properly with TrueGrid SecureRDP. TrueGrid eliminates firewall exposure and internet latency and provides secure remote access to your windows, desktops, and apps. So check out all of them.

You can head over to itbusinesspodcast.com slash sponsors. Folks, please support my folks. They are doing some great stuff.

And of course you can support the show by clicking on the Amazon link on the website. I want to thank you all again. We are doing fantastic for the second month in a row.

All you have to do is click there first and anything that you buy, same price, but Amazon takes care of stuff in the back and kicks me back a little, one, two percent. And it adds up and helps me pay for all the things that we do here for the show. And let's see, any other housekeeping stuff?

I think we're good there. I mentioned all of that. Got the code, TechCon Unplugged, be there.

There is a pre-day for TechCon Unplugged. I didn't mention that for whatever reason. Super Ops is doing a pre-day event.

Let's see. I have nothing in my notes. I know they're doing something; space is limited.

Priority given to the first 50 registrants. I don't even know if they're full or not. So if you try to sign up for that and they're full, sorry.

But at least I mentioned it there. All right, Rob, let's see here. We talked a lot of crap about AI and contracts, but I think if we were to talk about what the best practice should be, of course, the first thing is whatever you do, it needs to be taken to your attorney for review, correct?

[Rob Scott]
Yeah, I think that when it comes to best practices involving risk, there's three important components. One is better contracts, and that's the point you're making. Two is the right insurance both for the provider and the end user customer.

And three is auditable controls by the provider so that if anything happens, we're going to have all the documentation, all the ammunition that we're going to need to defend a case in court. And so really operational maturity around security and compliance is critical. Insurance both for the provider and the end user.

And then, of course, contracts that make it clear that the MSP is not responsible for the criminal acts of third parties. Think ransomware, think business email compromise, and is not responsible for the acts or omissions of third-party service providers. Think Acronis in Sacramento, CrowdStrike outage, Kaseya ransomware attacks.

Anytime a vendor act or omission injures your customer, it raises the question under what circumstances could you be held liable? That's that whole duty of care issue we talked about earlier. And so those three categories are really where we focus our recommendations around best practices for risk mitigation.

[Uncle Marv]
All right. If somebody were to start with an AI generation, how important is it to keep detailed records of the prompts used, the outputs that you've gotten, the changes and stuff like that? Is that something that we should do before we bring something to you for review?

Or should we just let you draft it?

[Rob Scott]
Yeah, it depends on what it is. I'm happy to take from my clients any AI-generated content and think of it as a discovery call. Think of it in this context, Marv.

Instead of a replacement for a lawyer, what if everything you did with the AI was just to help me get up to speed on what's going on with what you're working on? And then I could take it from there. And therefore, we could leverage the power of the AI, reduce the time that we have to interact.

And now you've taken it as far as you can, but you had smartly included a lawyer in the loop who could take it that last mile and put their stamp on it and stand behind it. And so I think there are ways, for example, to do both, where you can use the AI and I can review it or someone else can review it. And there's a lawyer in the loop, but we're together leveraging the AI in a way that's different than let me use the AI to generate something as an alternative to using a lawyer.

What about if we looked at AI as a way to change the way the lawyer and the client communicate on the front end and maybe reduce some of the preliminary drafting where the lawyer is just adding that secret sauce at the end? So that's an example of how the client can benefit from using the AI without taking the undue risk of putting the lawyer out of the loop. So I think that's an important way to think about it.

I also think that when it comes to the questions that you asked about retaining session reports and outputs, I see that more important in the context of anything that you want to get protected, anything you want to own. If you want to get a copyright on something that contains AI-generated content, it's critically important to do what you said, which is you're going to need to be able to demonstrate the portions that were AI-generated and the portions that were human-created because that will come up in the context of the ability to own the outputs. The current state of the law in the US is nobody can own the AI-generated output.

The only question is, at what point does a work that contains both AI-generated content and human-created content subject to copyright registration and therefore ownership? And so in that context, it's critically important to retain the history because it's very relevant to the legal issue in copyright. In the context of using AI as a do-it-yourself tool for customer contracting, I don't think you have the same concerns.

The bigger concern there is what we've been talking about, which is if you don't have a lawyer in the loop and the AI doesn't get it right and you have a problem, now all of a sudden, what will happen? What will the recourse be? Will your carrier even cover you in a claim that's based on a contract like that?

These are all the things that are starting to happen, Mark. I mean, over time, what I think you're going to see in the managed services industry and what I mean by over time is within the next 12 months, most of the carriers will be requiring MSPs to use very specific types of agreements that the carriers will be dictating what those customer contracts need to look like. And the way they're going to enforce it is, if you use a contract that's different than the one they've approved, they're going to deny your claims.

So you start using AI-generated agreements because you want to save a few bucks, that could cost you your insurance in terms of coverage. And that's happening right now. These insurance carriers are coming out with rules that are dictating that the agreements need to be industry standard, industry specific with a broad array of legal protections, because after all, it's the carriers that are writing the checks, right?

[Uncle Marv]
Well, I was going to ask a question about that, because they're also doing the same things when it comes to the stack of protection in their cyber liability policies, where they're now offering up a suggested list of tools to use, antivirus, MDR, EDR, firewall protections. They're actually coming up with their own list of suggested stack items. And it seems as though they are going to kind of ring fence, here are the things that we will allow in our sphere of influence when it comes to granting or denying claims.

And if you're saying that they're also going to do that with contracts, it's really going to, I mean, in some sense, it can make our jobs easier. We'll just do whatever the insurance people say. But in other ways, it makes it harder, because what if we don't like the stack?

What if we don't, the contract doesn't really benefit us? I mean, where do we draw the line, first of all? And second, who is helping them draft these things?

[Rob Scott]
Yeah, that's an interesting question all the way around. We have partnered with TechRug, the MSP Alliance Insurance Program, and Lloyd's of London specifically, because all of those organizations are committed to a principle that's different than what you're talking about in terms of dictating the specific stack. These programs are designed to help MSPs mitigate risk, both professional liability and to help their customers get first party insurance.

But these carriers specifically, philosophically, are not getting involved in the deployment of any tools. They are asking questions about the tools that are provided. They're asking questions about whether or not there's MFA, but they're not dictating tools to be run.

They're not asking for any data back from any client sites. And that's just the model that I feel most comfortable with. And the one that I think protects the MSPs the most, because really, if the carriers are getting into managed services, then there's no real basis for a managed service provider to want to do business with a carrier that's just going to go behind their back and take their customers or offer solutions to their customers that are competitive with their stack.

It just doesn't align well. And that's why I think these vertically focused industry and community focused programs, like that run by TechRug and the MSP Alliance, are never going to go in that direction that you're talking about, because the people that are behind those organizations are philosophically against what they see those carriers doing to MSPs. And there's a different approach and mindset where you can get good insurance and never have to worry that the carrier is going to be competing with you on tools or dictating to you what tools have to be run or asking you to share any information specifically from your client environments, other than what would be in a questionnaire as far as an application is concerned.

[Uncle Marv]
All right. Well, that sounds good. Stay with those providers, so that'll work for us.

All right, Rob. I think I've asked all the questions that I could come up with in terms of using AI to write our contracts. And of course, I titled this podcast, Who's Writing Your Contracts?

Because in a sense, a lot of times it's people just going out there and generate a MSA, but where's it coming from? But I think what you said is probably going to be apropos, where we just need to make sure that whatever we do, whether we start with AI and then take it to an attorney, or we just go to the attorney. Having it reviewed is probably the best course of action, regardless.

[Rob Scott]
Yeah, for sure. And I think that the AI solutions offer a lot. For example, your service descriptions.

I'm less concerned about you using an AI tool to generate service descriptions that describe your business and have those reviewed by your lawyer periodically than I am you using a tool to write a legal agreement on the indemnity, limitation of liability, and insurance provisions of an MSA. Because you don't really know what's important in those provisions likely. Whereas your service descriptions are yours.

You're the subject matter expert on that. I feel very comfortable that you can use the AI tool to do that. That's the distinction between where to focus your energy on leveraging AI.

Leverage it in the areas that are in your domain. Use that to share with your lawyer so that they can get a head start. Don't think about AI as an alternative to lawyers.

Because if you do that, you're going to wind up in a bad spot with a bad agreement and nobody to blame but yourself.

[Uncle Marv]
Sounds good. Rob, thank you very much. Normally, this would be the point in time where I would say for you to, you know, either challenge a Florida man story or answer a random question.

But I'm going to let you off the hook. Part of the reason I want to do that is you actually, I was expecting you to make a comment about the upcoming, the Pitch It competition.

[Rob Scott]
I'm glad you brought that up.

[Uncle Marv]
For those that don't know, I actually, this year, for the first time, put together an Uncle Marv's top 10. I have a very proprietary secret formula that I use to generate the odds of the Pitch It contestants this year. And I came up with the top 10.

Rob, unfortunately, you guys weren't in there. And I think you took a little offense to that.

[Rob Scott]
Well, Marv, I just think that at least for the steak knives, Monjur has to be in contention. It would be a great honor for me to win Uncle Marv's steak knives, even though that's third place. You know, I saw a lot of people at the Olympics celebrating when they won bronze.

[Uncle Marv]
Yeah.

[Rob Scott]
And to be the only legal service in the group, to be a bootstrap, you know, self-funded company growing at 3x, adding 35 new logos a month, and helping a lot of MSPs sleep better at night. I think Monjur deserves at least a shot at the third spot.

[Uncle Marv]
Well, I will say this. The list is for entertainment purposes only. It was created without any professional guidance.

The rankings and opinions expressed are subjective and may not reflect actual data or insights. And it obviously does not reflect who I know and like in the industry. So that's my disclaimer there.

[Rob Scott]
Hey, listen, if I was ConnectWise, I probably wouldn't want Monjur to win.

[Uncle Marv]
Uh-oh.

[Rob Scott]
And so because of that, I will say to you that I think that you're right. The rankings are probably pretty accurate on first place, but I just want to win those state medals.

[Uncle Marv]
All right. All right. We shall see.

And we'll find out in November. I'll have them with me and make the presentation there.

[Rob Scott]
I would be honored to receive them from you.

[Uncle Marv]
Okay. All right. Let me quickly review two things about Florida Man.

I've already gotten a couple of questions because this past Sunday, the 2024 Florida Python Challenge was completed. And last year around this time, I actually announced the winning Python in terms of length and size. And people were already asking about that.

The results have not yet been released. But I can tell you that this was a very successful season. Over 600 people registered for this year's challenge, including 108 participants from other states and two came down from Canada.

Last year's total was a record 209 pythons captured. And of course, the goal is to surpass that. And the grand prize winner will receive $10,000.

So that's where we are. As soon as those results are released, I will announce them here. Don't know why people are so interested in the Python Challenge, but that's it.

And then the Florida Man story for tonight. A 55-year-old man named Jose Marti Alvarez has been arrested for impersonating Romans Pizzeria in Miami Springs, Florida. He was running a fraudulent scheme that misled tourists into purchasing substandard pizzas.

Marti Alvarez allegedly operated this scheme for several years, distributing counterfeit flyers in nearby hotels that falsely advertised his business as Romans Pizzeria. And customers receiving the pizzas reported getting them undercooked, incorrect or fraudulently priced, leading to numerous complaints against the legitimate establishment and damaging its reputation. So I will have the full story there.

So Jose was arrested and has been released on a $5,000 bond. Despite the negative impact on his business, Roman, the real business, indicated that he did not seek revenge, but rather hoped that Jose Marti Alvarez would stop using his restaurant's name and operate under his own and serve quality food instead. So there is your Florida Man story for tonight.

So I'll have the link there. And I apologize, last week I did not have the Florida story link in there that was mentioning Florida as the number two state in the country, best place to live, behind Massachusetts. So I'll get that link up as well.

You guys can read the full list there and enjoy everything. That's going to do it for tonight. Next week, Marti Stockman will be on the show.

I've already released part one of an audio show that I did with her. That went out this morning. Part two, I believe, is scheduled for Friday.

So you can listen to those two episodes and then join us live next week. And we'll chat about all things Marti-related, Ted Lasso-related. Et al.

And then, of course, Rob, we'll see you in the Pitch It competition. When is your... You didn't go today, did you?

[Rob Scott]
I did Battle Royale this week.

[Uncle Marv]
Are you on Thursday?

[Rob Scott]
No, I already went.

[Uncle Marv]
Oh, you went yesterday.

[Rob Scott]
Yeah, we were day one. A lot of really great companies in the competition this year. Thought our presentation went very well.

One of the things they ask you about is your likelihood of success as a company. And I don't think there's any question about Monjur’s likelihood of success with 3x growth, over 450 MSPs since March of 2020, and growing at 35 new logos a week. I think Monjur’s here to stay.

[Uncle Marv]
All right. Sounds great. Congratulations on all that success so far.

Wish you luck in the Pitch It contest. We'll see you in Orlando. And thank you for coming on the show tonight.

[Rob Scott]
Looking forward to seeing you at TechCon as well, Marvin. Thank you for having me, as always. It's great to see you, and I appreciate your hospitality.

[Uncle Marv]
I appreciate you too. All right, folks, that's going to do it for tonight. Thank you very much for downloading and subscribing all of our shows.

Head over to IDBusinessPodcast.com. Check out other shows, support the sponsors, shop on Amazon. That's going to do it for tonight.

We'll see you next week. And until next time, Holla!