Latest / The Insurance Marketer / How AI is Reshaping Marketing Communications w/ Sharon Toerek
Transcript
- 0:00SPEAKER_02: From the OneFire studio, welcome to the IMCA Peer-to-Peer Podcast, where we dive deep into technology, brand, social media, and other topics with an insurance industry expert.
- 0:13SPEAKER_02: Let's welcome today's guest.
- 0:15SPEAKER_01: Hi there, I'm Jim Flynn from Onefire, along with my co-host, Leslie Castillo from America Equity Underwriters.
- 0:22SPEAKER_01: And it's our pleasure to welcome you to today's episode of the IMCA Beer to Beer Podcast, where today we'll be talking about how AI is reshaping marketing communications, where the most significant opportunities may lie, and where the most prominent legal challenges may be hiding.
- 0:40SPEAKER_01: And to help us better understand that journey, our guest today, Sharon Torque, is an international property and marketing attorney in Cleveland, Ohio, where she heads Torque Law.
- 0:50SPEAKER_01: Her practice works with marketing and creative professionals across the U.S.
- 0:55SPEAKER_01: to help them protect, enforce, and monetize creative assets.
- 0:59SPEAKER_01: And we all need to do that on a regular basis, right?
- 1:03SPEAKER_04: That's the hope.
- 1:04SPEAKER_04: That's the hope.
- 1:08SPEAKER_01: The marketing job, right?
- 1:09SPEAKER_01: And uh you guys, you're backing us up.
- 1:12SPEAKER_01: So she uh in her in her work, she counsels clients on marketing and creative legal issues, including copyright and content protection, licensing of creative content, trademark and brand protection matters, marketing contract issue, freelance contract issues, social media, advertising compliance, and direct marketing marketing regulations.
- 1:33SPEAKER_01: Sharon is the host of the Innovative Agency Podcast and the Popular LNC blog on her website at legal landcreative.com.
- 1:43SPEAKER_01: The blog discusses creative issues with a legal focus.
- 1:47SPEAKER_01: Sharon, you have honored us before uh uh being part of an IMCA webcast as our guest, and you're back by popular demand to discuss how technologies like AI, which were far from mainstream just a few short years ago, have have really changed uh critical legal issues.
- 2:04SPEAKER_01: So welcome back, uh Sharon.
- 2:06SPEAKER_01: It's great to have you.
- 2:07SPEAKER_04: Thank you, and thank you for having me back.
- 2:09SPEAKER_04: And it's um it's surreal to think that this is probably a topic that was not even on our radar the last time we were all together to have a conversation.
- 2:20SPEAKER_04: So I'm thrilled to be back and I'm looking forward to talking with you and Leslie both.
- 2:25SPEAKER_05: Yeah, the world is changing rapidly.
- 2:28SPEAKER_05: So we're very happy to have you uh walk us through this and avoid uh all the the minefield that is uh AI legal thing.
- 2:36SPEAKER_05: It's my pleasure.
- 2:37SPEAKER_05: Well, um Sharon, Jim mentioned that you're in Cleveland, Ohio.
- 2:41SPEAKER_04: I am lifelong Cleveland, Ohio girl.
- 2:44SPEAKER_04: Okay.
- 2:44SPEAKER_05: I was gonna ask where home is for you originally.
- 2:46SPEAKER_05: So you're born and bringing it.
- 2:48SPEAKER_04: I've lived in Cleveland my whole life.
- 2:49SPEAKER_04: We're a multi-state firm.
- 2:51SPEAKER_04: Um, and so we have people in multiple states, but I have always lived in Northeast Ohio.
- 2:57SPEAKER_04: My husband's from Northeast Ohio, and so we uh our families are both um from here.
- 3:03SPEAKER_04: And yeah, I kind of um built my business and my career um from what has been a great launching pad.
- 3:11SPEAKER_04: It's a great town to do business um nationally from.
- 3:14SPEAKER_01: So, you know, you've you've seen the lay of the land in legal issues from uh marketing communications for a while.
- 3:24SPEAKER_01: And as you said as we started here, AI has brought in a whole new set of set of problems.
- 3:31SPEAKER_01: So I I mean, two-part question.
- 3:34SPEAKER_01: The first part of it, you know, where you know, where where are you seeing clients using AR to really reshape the way that they're working?
- 3:45SPEAKER_01: And, you know, where do you see the biggest minefields that they're running into right off of the bat?
- 3:52SPEAKER_04: Yeah, it's interesting, you know, when um sort of the conversation around Chat GPT sort of democratized talking about AI.
- 4:02SPEAKER_04: The truth is we've been using AI in our businesses for a long, long time, right?
- 4:06SPEAKER_04: But the the buzz got loud when Chat GPT became more of a, shall we say, retail product?
- 4:15SPEAKER_04: And I would say the first thing that marketers, particularly on the agency side or the marketing services provider side, and and brands thought about this too, was what is this gonna mean for our workflow, our processes?
- 4:30SPEAKER_04: Um, what's it gonna mean for how humans actually participate in creating the deliverables and developing strategy?
- 4:37SPEAKER_04: What's it gonna mean for our jobs?
- 4:39SPEAKER_04: So honestly, we spent a lot of time, our clients, I should say, spent a lot of time worrying about that initially.
- 4:46SPEAKER_04: And then that kind of settled down a little bit as it kind of began to become obvious that the opportunities were gonna outweigh for most of us.
- 4:54SPEAKER_04: I mean, there will be some marketers who choose not to adapt or won't be able to adapt to using the tools and integrate them who might fall behind.
- 5:02SPEAKER_04: But for most of us, we're seeing the possibilities of integrating AI solutions into our processes behind the scenes to make the marketing more effective, to make it um more insightful, to get better results, better data, and also to drive down the cost, quite frankly, which is I think what a lot of us were worried about.
- 5:21SPEAKER_04: And then along with that, and in answer to the second part of your question, I would say the initial concerns that were sort of bandied about, and still we don't have answers to these questions.
- 5:31SPEAKER_04: We have somewhat of a blueprint, but who is gonna own all of the stuff that we use generative AI in particular to create?
- 5:42SPEAKER_04: Number one.
- 5:43SPEAKER_04: And number two, what are the conversations that brand marketers and their agencies need to be having about that?
- 5:51SPEAKER_04: What are the expectations?
- 5:53SPEAKER_04: How do we manage those expectations?
- 5:55SPEAKER_04: How do we not take actions that are going to get our companies, or if you're an agency, our clients sued, you know, for copyright infringement?
- 6:02SPEAKER_04: Um, and what are some of the other landmines like data, access to data, and um inputting of confidential data into some of the AI platforms?
- 6:12SPEAKER_04: So that's kind of what I think is driving most of the concern and questions from a legal perspective around integrating AI into marketing work.
- 6:22SPEAKER_05: As you know, many of our listeners market property and casualty insurance products to B2B companies.
- 6:29SPEAKER_05: And they are uh increasingly using AI-based tools to create content, automate digital ad production, plan media schedules, and you know, various other tasks, uh, anything that helps us do do more with less, like we're we're all over it.
- 6:45SPEAKER_05: So, what would you say are the primary risks associated with using AI to develop marketing strategies or these various tools?
- 6:54SPEAKER_05: I mean, particularly in a highly regulated industry like insurance.
- 6:59SPEAKER_04: I think there's a couple points of particular vulnerability when you're in uh one or two of them might be specific to being in a highly regulated industry, but you know, the others would be applicable even if you were not.
- 7:11SPEAKER_04: First of all, how much do they care about owning the end creative product?
- 7:16SPEAKER_04: Because the law is very clear right now, as we're sitting here at the beginning of 2025, that nothing generated from the machine um is really ownable from a copyright perspective.
- 7:29SPEAKER_04: What's more of an open question is what happens if the output from an AI platform happens to infringe upon somebody else's actually copyrighted work.
- 7:38SPEAKER_04: And so there's lots of you know debate about what that's gonna mean and what the liability where the liability is gonna lie.
- 7:44SPEAKER_04: So that's the first thing, and that would apply, I think, regardless of what industry um that we were in.
- 7:50SPEAKER_04: But more specifically, given the fact that insurance, finance, healthcare, highly regulated industries involving lots of personal information, um, which is necessary sort of to make marketing effective, is vulnerable to data breaching when they're using these AI platforms to sort of um, because you inevitably have to feed them in, you know, they're only as good on the output side as what you put in to either train the models or feed the beast.
- 8:21SPEAKER_04: Yeah, I mean, right?
- 8:22SPEAKER_04: Garbage in, garbage out.
- 8:23SPEAKER_04: And so nobody wants garbage.
- 8:24SPEAKER_04: And so you're putting in data.
- 8:27SPEAKER_04: And so from a regulatory perspective, what are your obligations to make sure you're not breaching any of your patient or client or customer confidentiality, you know, rights?
- 8:39SPEAKER_04: And also, what sort of inputs are you creating to use AI that might include strategically sensitive information?
- 8:50SPEAKER_04: A lot of marketers and the agencies and vendors they employ are having great success getting um market research done, creating strategy and market insight documents, using AI tools, but you have to input data into those to actually to get out a product that's going to be useful to you.
- 9:08SPEAKER_04: And so, where are the dividing lines and how much information are you putting at risk by feeding this data into your prompts and into these tools?
- 9:20SPEAKER_04: Um, and what will be the outcome of that?
- 9:22SPEAKER_04: And so the higher the level of regulatory risk in general for an industry, and obviously for insurance and financial services, it's quite high, but it's not zero for any industry.
- 9:35SPEAKER_04: So I think those are the two main um legal issues that from a regulatory or a legal perspective, brands and their agencies have on their mind.
- 9:45SPEAKER_04: Then there's all sorts of sub-issues that, you know, um sprout up around that, like how are we going to design our contracts so that we are putting the responsibilities in the right place?
- 9:55SPEAKER_04: How are we handling our communications either internally as a company or with our outside partners to make sure that everybody's on the same page about how and when AI is going to be used, what platforms are allowable to use, and what are the use cases that are going to be permissible, and what are the policies we need to write and socialize within our organization so that everybody understands that?
- 10:22SPEAKER_01: It sounds like much of this is still in discussion and ultimately will end up in case law that will end up driving all of this.
- 10:34SPEAKER_01: What are when you're receiving calls, what are the biggest misconceptions about data privacy and the risks that that people are seeing out there when they're leveraging uh AI?
- 10:49SPEAKER_01: I mean, you must get a lot of, well, I didn't realize that.
- 10:52SPEAKER_01: Well, I didn't know that.
- 10:54SPEAKER_04: Yeah, uh we do.
- 10:55SPEAKER_04: And I think it's less we do.
- 10:57SPEAKER_04: I think it's less of a misconception than a lack of awareness, honestly, or visibility to the issue.
- 11:03SPEAKER_04: I don't think that the users of these platforms are necessarily thinking at a global level about the implications of the data they're inputting into the machine and what that could mean on the other side of it.
- 11:19SPEAKER_04: Might that data show up in somebody else's output?
- 11:21SPEAKER_04: And we didn't intend that, or might we inadvertently be breaching, you know, privacy rights of some of our customers or patients by using this AI platform in this way to create this result.
- 11:35SPEAKER_04: And so you really, it's a multidisciplinary approach in any insurance or financial services organization between legal, marketing, IT, really it's a multi-stakeholder conversation that has to occur.
- 11:50SPEAKER_04: Um, and this is why, quite frankly, and I, you know, my apologies to the brand side folks that are listening to this, but this is why, you know, agencies right now, at least as we're sitting here today, um, are moving faster with their adoption of AI because they don't have as much of this baggage really dragging them down.
- 12:09SPEAKER_04: And they're faster to create their own policies and and to be more, you know, dynamic about how they adjust and change them.
- 12:16SPEAKER_04: In a big brand culture, it's very hard to move quickly enough.
- 12:20SPEAKER_04: And so you have this balance between seizing opportunity and wanting to use these tools for the efficiencies that Leslie was mentioning earlier, with being very risk averse, which you're in the insurance industry.
- 12:32SPEAKER_04: You've got one of the most risk-averse group of people in business who I know.
- 12:37SPEAKER_04: And so it's, you know, it's a constant high wire and an intention dance between those two things.
- 12:44SPEAKER_05: You mentioned earlier that it's it's becoming very clear that the benefits of AI outweigh the cons.
- 12:52SPEAKER_05: And one of the cons of the content that's generated from AI, at least from my perspective, I think Jim would agree, is that it sometimes feels like it lacks a certain humanness that's intrinsic to many aspects of insurance, right?
- 13:07SPEAKER_05: Like we always say this is a people business, but yet we're starting to use tools that generate content that don't always sound like a person wrote it.
- 13:15SPEAKER_05: So from your perspective, you know, legal or otherwise, how can companies in the insurance industry balance the benefits that are provided by AI automation with the need for a more personalized, human-centered customer experience?
- 13:30SPEAKER_04: You know, I do think that is more of a marketing than a legal question, but I will say this based on my experience in talking to marketers just generally.
- 13:38SPEAKER_04: I think this is just an example of a point that's really important to keep at the forefront for both legal and marketing purposes, which is that you have to take a position about the human guardrails you're going to put around all of this.
- 13:52SPEAKER_04: Using AI tools is not a substitution for human input.
- 13:57SPEAKER_04: And whether it's in crafting the messaging or in um industry insights or knowledge that the that the are a function of institutional experience that you know humans have working in your industry have.
- 14:12SPEAKER_04: Um, those are the same kind of guardrails that really matter in terms of reducing risk in um putting the product out there that AI has helped create.
- 14:22SPEAKER_04: I do think we're already seeing a lot of mediocre, particularly when it comes to direct response marketing.
- 14:29SPEAKER_04: Um, I think we can all kind of tell when we're getting emails now and AI has written them.
- 14:34SPEAKER_04: You certainly can tell when someone is reaching out to us on LinkedIn and AI wrote, you know, that message.
- 14:40SPEAKER_04: So, I mean, marketers are getting very savvy, I think, about being able to determine when there wasn't enough human insight put into the strategy, the messaging, whatever it is.
- 14:52SPEAKER_04: Same thing for the legal, you know, due diligence required.
- 14:55SPEAKER_04: It really requires the brand to think carefully about what are the human guardrails and inputs we are going to require around all of this stuff.
- 15:03SPEAKER_04: How many eyeballs need to look at this and whose eyeballs do they need to be?
- 15:07SPEAKER_04: And that I think Leslie solves both problems because it prevents the sort of mediocrity or the lack of insight that you want to avoid when you're spending money putting marketing out there.
- 15:19SPEAKER_04: Um, and also it's another guardrail against legal risk because um some, you know, someone else has taken a look at it too, this is wrong, or if we input this data, this might be the outcome of that, and that's a problem, um, et cetera.
- 15:36SPEAKER_04: For sure.
- 15:36SPEAKER_01: So there's a lot of uh legal potholes, it sounds like.
- 15:40SPEAKER_01: Um and and some ethical ones as well, which um I mean, I guess uh eventually an ethical issue will turn into a legal one, will it not?
- 15:53SPEAKER_01: Yeah, generally, generally will.
- 15:55SPEAKER_01: Right.
- 15:56SPEAKER_01: Um, so uh boy, and as you're looking at using AI to model customer behavior, you talked about that a little bit.
- 16:07SPEAKER_01: And uh it getting ending up in you know in in an unintended place.
- 16:13SPEAKER_01: Can you give us kind of some background on what that looks like and and how it becomes um a problem for the marketer?
- 16:25SPEAKER_04: Um, this is an extreme example, and I'm making it up on the fly, but let's say you want to create um a brief that contains strategic insights about a specific segment of the insured community, and um, you know, you are feeding into the platform of your choice um a prompt that asks it to give you insights for this particular uh group of clients having this particular characteristics, um, and you are actually feeding in data that's not anonymized properly, or it's inevitable, you know, in uh indiscreetly and accidentally um inputting personal data of kinds and customers, that's a problem because those that data is in the system, if you will.
- 17:18SPEAKER_04: Now, many platforms have multiple layers of access wherein they claim they're not using data in prompts to train um their models.
- 17:30SPEAKER_04: You don't know until you know whether there's a problem or not, is my answer to that.
- 17:34SPEAKER_05: That feels scary.
- 17:36SPEAKER_04: Yeah, well, I mean, you have to decide really how repeat more the vendor you're dealing with is and how much you can accept their representations and what the risk would be to you if the data inadvertently got breached.
- 17:50SPEAKER_04: The other thing is if you are asking for assistance from AI to help create strategy ideas, how much of your secret sauce do you have to include in the inputs to get a meaningful recommendation?
- 18:07SPEAKER_04: Um, and again, these are business and marketing questions, not really legal questions, but the legal aspect of it is what confidential stuff are we feeding in on the input side in order to get output that's going to be useful to us?
- 18:20SPEAKER_04: And you have to think about that every single time you use it.
- 18:24SPEAKER_04: And you have to trust that you have a team that has been educated about the right questions to ask, or that you've got appropriate guardrails, you know, managing that that that function so that you're not making mistakes that um for lack of oversight.
- 18:42SPEAKER_01: And it really doesn't, I mean, it it doesn't have to be um personal data that you're in putting.
- 18:49SPEAKER_01: I mean, it it can really be secret sauce kind of data.
- 18:53SPEAKER_04: Yeah.
- 18:53SPEAKER_01: I'm about categories and so forth.
- 18:56SPEAKER_04: Totally.
- 18:57SPEAKER_04: And I think that that is far more likely to be a risk for most companies that I think most companies intuitively know, let's not put our customer list into a prompt and ask for insights based on their characteristics.
- 19:08SPEAKER_04: That's just, I think, commercial suicide, if you will, because that's an that would be that would have been illegal before AI was so widely, you know, deployed in marketing functions.
- 19:18SPEAKER_04: But it may not be so obvious that um either somebody internally to the organization or with one of your marketing vendors or a contractor working for one of your marketing vendors just you know fed the data in um and asked for some starting points for a strategic brief or um marketing copy based on this message, based on this new product release.
- 19:47SPEAKER_04: You know, and those are the things that once that data's in there, it's in there.
- 19:51SPEAKER_04: And so um what happens to it is, you know, kind of out of your control.
- 19:57SPEAKER_05: Well, and I do think that's the bigger risk that is.
- 20:00SPEAKER_05: You've got me thinking that people in marketing function, so marketing teams are often um, I don't know if privy is the right word, but like privy to secret sauce that they might not realize is secret sauce, like because it might just be like this is our strategy and these are our objectives, and they might not realize um, especially at at a more junior level, um, some marketers might not realize that some things are actually strategic uh things that might not, you know, you might not want to put in Chat GPT.
- 20:34SPEAKER_05: Yeah.
- 20:36SPEAKER_01: We'll take a quick break and come right back.
- 20:39SPEAKER_00: Are you looking to take your company's growth to the next level?
- 20:42SPEAKER_00: Look no further than OneFire, the RevOps-focused marketing agency that specializes in helping insurance companies grow.
- 20:49SPEAKER_00: Our expert team of strategists, marketers, and HubSpot specialists will help you optimize your marketing efforts and maximize your RLI.
- 20:57SPEAKER_00: From lead generation to conversion and retention, OneFire has you covered with our deep insurance industry expertise and modern approach to growth.
- 21:06SPEAKER_00: Don't let your competitors leave you in the dust.
- 21:09SPEAKER_00: Ignite your growth with Onefire.
- 21:12SPEAKER_00: Visit us today at onefire.com.
- 21:16SPEAKER_05: You had mentioned that the industry is inherently risk averse, and you're right, but you know, we're speaking of risky behavior.
- 21:24SPEAKER_05: So what steps should marketers take to mitigate the risk of um AI-driven bias or data misuse?
- 21:34SPEAKER_04: I think it starts with crucial conversations.
- 21:37SPEAKER_04: First of all, what is the goal we're trying to achieve with this activity that we need to conduct?
- 21:44SPEAKER_04: How do we think we can best use an AI tool to help us achieve that goal?
- 21:50SPEAKER_04: What what we have to do in order to um activate the AI to help us here?
- 21:58SPEAKER_04: And what What and who's going to be involved?
- 22:01SPEAKER_04: I think if you start with those questions, um, and then you take a step back and think about what is our company's policy and risk tolerance level for AI overall?
- 22:12SPEAKER_04: And everybody listening to this, their organization should have a viewpoint on it.
- 22:17SPEAKER_04: Um, and that viewpoint should be expressed in written policies that are written in plain English, that everybody internally is socialized in, and that your vendors and strategic partners are aware of.
- 22:30SPEAKER_04: So, what are our policies and our risk tolerance overall around using AI?
- 22:35SPEAKER_04: What might be at odds between our policies and our goal here for this project?
- 22:42SPEAKER_04: And let's think it through.
- 22:44SPEAKER_04: So the challenge is that employing all those steps, um, I'm not living in a dreamland here because I represent marketers every single day, and you all want to work fast.
- 22:54SPEAKER_04: This is not fast friendly necessarily.
- 22:57SPEAKER_04: And so we don't want to lose all the economy of using AI having to double and overthink everything.
- 23:07SPEAKER_04: The point is if you've got the infrastructure set up in the first place to have the right conversations, to go through the right checkpoints, it shouldn't slow you down.
- 23:16SPEAKER_04: Um, but it starts with the organization thinking at a high level about what is our position here and have we adequately communicated it internally and to the people we hire to help us do our marketing.
- 23:30SPEAKER_04: Because I can promise you that it's not a conversation that brands are always leading.
- 23:35SPEAKER_04: And um, some agencies will lead the conversation, but not all of them.
- 23:40SPEAKER_04: So if neither of you are leading it, then it doesn't come up until a specific project with a deadline attached to it is on fire.
- 23:47SPEAKER_04: And then it's kind of late to sort of slow yourselves down.
- 23:51SPEAKER_01: No, absolutely.
- 23:53SPEAKER_01: You know, what what do you see on the the immediate horizon, I guess, short term and and longer term?
- 24:00SPEAKER_01: Um, you know, what changes out there legally should insurance marketers really anticipate happening?
- 24:06SPEAKER_01: What what regulations are gonna, what case law is gonna drive this kind of stuff?
- 24:12SPEAKER_04: I think that the best shot we have at seeing any advance in case law um will be in the IP arena.
- 24:20SPEAKER_04: There are um a number of copyright cases in litigation right now, you know, lodged on behodged against um OpenAI, which is um ChatGPT's parent company, um, regarding the um inputting of definitely copyright-owned content into ChatGPT and whether or not that is an infringement of the copyrights of the original creators of that work.
- 24:48SPEAKER_04: OpenAI has settled with the Associated Press, um, with a licensing arrangement, uh, but the New York Times case is still very much alive, and we're waiting to see how that will progress in terms of will it be considered an infringement of copyright to actually even input this stuff into the um the AI machine for data training purposes, or is that fair use?
- 25:17SPEAKER_04: And if it is, um, is using the output, even if it's substantially similar to the copyright-owned work, is that a copyright infringement?
- 25:26SPEAKER_04: So I think we'll see some forward movement and a decision on are they gonna split the baby or are both those things gonna be copyright infringement?
- 25:34SPEAKER_04: We don't know yet.
- 25:35SPEAKER_04: I don't think we're gonna see um over the next, I mean, as we're sitting here in early 2025, I do not expect to see significant regulatory clarity on data privacy.
- 25:47SPEAKER_04: It we don't even have a national data privacy standard um in the United States as it is.
- 25:55SPEAKER_04: I don't expect um AI to I mean, we're hopeful that the issues that AI raises might light a fire under legislators to do something federally, but I don't think we're in an environment for that to happen right now.
- 26:08SPEAKER_04: So I don't expect anything significant on the data privacy front.
- 26:12SPEAKER_04: We're gonna have to rely upon the patchwork of legislation that we're already working under GDPR, the California laws, which are the most sort of gold standards you can apply to things in the United States.
- 26:25SPEAKER_04: Um and so I don't expect a lot of activity there.
- 26:29SPEAKER_04: And then just hand-to-hand combat in terms of copyright infringement.
- 26:34SPEAKER_04: I think we're gonna have to see how much critical mass or volume we actually see in the terms of actual copyright owners seeing somebody else use their stuff who got it out of a prompt into um an AI platform.
- 26:49SPEAKER_04: So that's I think what we need to watch for on the case law and on the regulatory front.
- 26:56SPEAKER_04: And I think the case law stuff might move faster than anything regulatory.
- 27:00SPEAKER_05: While we don't have a national standard for data privacy, many companies in our industry do.
- 27:07SPEAKER_05: And if they don't, they're definitely thinking about it, right?
- 27:10SPEAKER_05: Um so as AI technology continues to advance rapidly, do you have recommendations for how companies in the insurance industry can future-proof their data privacy strategies?
- 27:24SPEAKER_04: I think the fact that you are in such a highly regulated industry actually gives you some cover that some of your peer industries don't have because AI will not enable you to do magically do some of the things that you're already prevented from doing in terms of um management and stewardship of that data.
- 27:45SPEAKER_04: Safeguard-wise, I think it's again having that multidisciplinary approach between IT and legal regarding how do we, what's our policies about where we house, who has access to, and who can deploy data outside the organization.
- 28:05SPEAKER_04: And I think that is the really the best first defense against making a mistake.
- 28:10SPEAKER_04: Um, I I think that as I think it's going to be challenging and interesting to see as the AI purveyors increasingly offer higher and higher levels of indemnification, liability protection, um, to see whether they will expand that to data privacy violations for users.
- 28:31SPEAKER_04: And if they do, and if those are legitimate indemnifications and liability limitations, um, then we might see a little bit more envelope pushing on behalf of the brands who want to use, uh who want to use AI to manipulate their data to create some sort of result to move the business further along.
- 28:50SPEAKER_04: So we'll have to see.
- 28:52SPEAKER_01: You mentioned the word uh socializing within an organization, you know, socializing the use or the regul, not necessarily the regulations, but the policies, so that, so that everyone understands what that looks like.
- 29:05SPEAKER_01: As much as it's possible at a company for everyone to understand anything.
- 29:09SPEAKER_03: Right.
- 29:10SPEAKER_01: You know.
- 29:11SPEAKER_01: Um so and you know, there's there's a lot of talk that I hear amongst our members of, you know, some some of them can't use AI at all.
- 29:24SPEAKER_01: I mean, some of them aren't allowed to log into Chat GPT.
- 29:28SPEAKER_01: Um and so what they're doing is they're doing it on their phone or they're doing it from you know at home, you know, so they can still play with the tool.
- 29:35SPEAKER_01: Um I didn't say that really out loud, right?
- 29:37SPEAKER_01: Did I?
- 29:38SPEAKER_01: Uh but yeah, but we know that that's happening.
- 29:41SPEAKER_01: And so a lot of times marketers, um, you know, they are at the you know not at the bottom of the food chain, but things get pushed down to them, you know, regulations too.
- 29:52SPEAKER_01: So from a proactive basis, what role should marketers and marketing teams play in helping shape the AI policies within their organizations?
- 30:04SPEAKER_04: Well, some of your listeners are going to hate this answer.
- 30:09SPEAKER_04: They need to be drivers of the conversation and they actually need to start the conversation of about documenting the standards and the expectations.
- 30:20SPEAKER_04: Uh, if that means the CMO or their designee um reaching out to their counterparts and the other parts of the organization to put together a task team to actually go into the organizations, um, you know, go deep in the organization and ask for input about how do you anticipate using these tools in your day-to-day work, or what opportunities do you want us to know about?
- 30:46SPEAKER_04: I think it starts with asking the questions of the um the users of these tools, assessing that data, and then in a cross-disciplinary way, developing policies that that can accommodate that as much as possible.
- 31:03SPEAKER_04: And not everybody's desired use is going to be um accommodatable, and nor should it be.
- 31:09SPEAKER_04: Um, again, marketers like to push envelopes.
- 31:11SPEAKER_04: I, you know, I live this world in this world every day, and so I know that.
- 31:15SPEAKER_04: Um, but uh as long as everybody's got a hand in creating the policy, the socialization of it is less challenging because they would have already been part of the building team.
- 31:28SPEAKER_04: Um and that maybe sound a little bit like utopia to some of the people listening.
- 31:33SPEAKER_04: But that would be an ideal world, that is the most organic way to go about as inclusive and realistic a policy as possible, um, is to really get input from the people on the ground floor of the organization.
- 31:48SPEAKER_04: And then making sure that um there's consensus amongst leadership about how AI can be used, what we need to say to our vendors about using it, who needs to look at agreements to make sure that that they are reflective of how we say.
- 32:04SPEAKER_04: Um because otherwise, the easiest thing to do, if organizations don't do that, the easiest thing to do is just knee-jerk no.
- 32:11SPEAKER_04: It's just knee-jerk legal tells everybody you can't use it, come see me if you have a question.
- 32:19SPEAKER_04: And that there's nothing wrong with that, but it's not conducive to integrating AI into the organization.
- 32:27SPEAKER_04: And it's not a sustainable long-term strategy either, because these tools are moving too quickly, and every vendor who every organization relies upon is using them.
- 32:38SPEAKER_04: And so, um, you know, kind of taking a we're not gonna go there viewpoint is a very short-term strategy, in my view.
- 32:47SPEAKER_05: The tools are moving too quickly, and so has our time together today.
- 32:52SPEAKER_05: Um, yeah, I can we could sit here and ask you so many more questions about A.
- 32:57SPEAKER_05: Great questions.
- 32:58SPEAKER_05: I love it.
- 32:59SPEAKER_05: Thank you.
- 32:59SPEAKER_05: Well, you've had great answers.
- 33:00SPEAKER_05: Um, I have no doubt that many of our listeners will have um you've given them a lot to think about.
- 33:07SPEAKER_05: And if they do have questions for you and would like to reach out to you, what would be the best way for them to do so?
- 33:13SPEAKER_04: Well, first of all, thanks for having me again.
- 33:15SPEAKER_04: And yes, I'd be happy to talk to your listeners.
- 33:19SPEAKER_04: It's um you can reach me on LinkedIn, it's Sharon Torik, T-O-E-R-E-K.
- 33:23SPEAKER_04: Um, and our firm uh email is legalandcreative.com.
- 33:28SPEAKER_04: So you can find us there.
- 33:30SPEAKER_04: And always happy to um, you know, to share ideas and expertise.
- 33:36SPEAKER_04: And we have a team that is kind of immersed in this every day for our agency clients.
- 33:42SPEAKER_04: And so we're always happy to talk about it and always happy to learn about new situations that require us to think about things differently.
- 33:49SPEAKER_04: That's always fun to do.
- 33:50SPEAKER_04: For sure.
- 33:51SPEAKER_01: Well, uh, yeah, I'll I'll say two things related to that.
- 33:54SPEAKER_01: One, um, yeah, your team is great.
- 33:57SPEAKER_01: And anytime I've ever had a had a question, they've been able to um to help very quickly.
- 34:04SPEAKER_01: And uh the other part of it is is um I I always find it very interesting when we've had a conversation and you say, Well, that's interesting.
- 34:17SPEAKER_01: I have that sinking feeling in my stomach.
- 34:20SPEAKER_01: Uh-oh, this isn't something that we talk about every day.
- 34:22SPEAKER_03: If that's what you want to hear from a lawyer, right?
- 34:25SPEAKER_03: You don't want to be using the word interesting.
- 34:27SPEAKER_01: Well, from your lawyer or your doctor, you don't want to hear that.
- 34:31SPEAKER_03: That's true.
- 34:32SPEAKER_03: I feel you there.
- 34:32SPEAKER_03: That's for sure.
- 34:34SPEAKER_03: Yeah.
- 34:34SPEAKER_01: Well, thank you so much.
- 34:35SPEAKER_01: That is a wrap on today's episode of the IMCA Peer-to-Peer Podcast, where uh we hope that you gain some valuable insights on how AI is reshaping marketing communications, where the biggest opportunities may lie, and more importantly, where the most significant legal challenges may be hiding.
- 34:55SPEAKER_01: If you uh enjoyed the show today, please uh please review our episode.
- 35:00SPEAKER_01: And don't forget to subscribe.
- 35:02SPEAKER_01: Give us a five-star rating, and if you do leave us a review, that will help us reach other listeners just like you.
- 35:09SPEAKER_01: And uh, if you'd like to share your thoughts on topics that you would like us to cover in future episodes, shoot them over to us at uh email at ideas at imca peer, the number two peer podcast.com.
- 35:24SPEAKER_01: Once again, this is Jim Flynn from One Fire, along with my co-host Leslie Castillo from American Equity Underwriters.
- 35:32SPEAKER_01: And it has been an honor and a pleasure to spend our time together today with our guest, yay, Sharon Tork.
- 35:40SPEAKER_01: Thanks so much again, Sharon, for being part of today's IMCA Peer to Peer Podcast.
- 35:46SPEAKER_02: Thanks to both of you.
- 35:46SPEAKER_02: My pleasure.
- 35:47SPEAKER_02: The Peer to Peer Podcast is a presentation of IMCA.
- 35:51SPEAKER_02: To learn more about today's episode and listen to other episodes, visit IMCANet.com.
- 35:56SPEAKER_02: This has been a production of One Fire.