The Marketing AI Pulse Brief for Feb 2026: Trust in the World of LLM Ads, OpenClaw, Reddit & More!

Welcome to the Marketing AI SparkCast, where AI meets modern marketing. Hosted by Abby Varma, a global leader in marketing, AI and innovation. Each episode brings you real conversations with leaders putting AI to work in strategy and execution. This episode is brought to you by SparkNovus and the Marketing AI Pulse community. Here's your host, Abi Varma. Hey, everybody. Welcome to the second episode of the Marketing AI Pulse monthly brief. This is where we curate stories that may be relevant for CMOs and marketing leaders to keep their finger on the pulse of marketing and AI. This is an extension of our Marketing AI Pulse community by the same name. If you are not familiar with the community, you can check out MarketingAIpulse dot com. And with me, I have the distinct pleasure of welcoming my partner in crime, the lovely, as always, Matt Seer. Matt, welcome to our second episode. Good morning, mister Varma. How are you today? I'm doing good, man. How are you? Sounds like you're a little under the weather and you're powering through. Yep. I have been fighting some some sort of a goblin here for the last week or so. But, yeah, can't you know, AI never sleeps, Abby, so we gotta keep on pushing. There you go. This episode, Matt, we have a lot of cool things to talk about. And one thing which is top of mind is advertising in the LLMs. You know, all these different LLMs have sort of a a distinct take on what that really means for their business. So ChatGPT, as you know, is all in on the free version and in the low cost version and are already testing advertising here in the US. Anthropic had those series of ads that they made. We're talking about not wanting to advertise, and now perplexity pulls back from advertising. So unpack this for us and tell us what it is and why that is relevant for marketing folks. Yeah. Definitely. It's been an interesting few months in this area. Right? OpenAI announced maybe two months ago that they were gonna start testing ads within platform. That brought some immediate backlash, not least of which were the Super Bowl ads that Anthropic released that you mentioned, which were, you know, if you saw them live or have seen them since, they were really well done and I think did a nice job of sort of creating some space between Anthropix approach to this and OpenAI's approach to this. And I think what you're seeing, especially now with this Perplexity announcement just in the last few days, is sort of an emergence of two different camps here when it comes to monetization. On the one hand, you have OpenAI and Google, which are going down the the reach path, which, you know, both of them are testing ads in their LLMs and trying to monetize in sort of, I would say, traditional ways. The other path, though, is what both Perplexity and Anthropic are doing, which is what I would put in the the trust category. Right? So you're seeing this emergence of a reach versus trust separation here. And I think what Perplexity is saying in their in their announcement about and we have to remember, Perplexity has had ads in context ads already. And what they're saying is we're actually now going to stop doing that and really just rely on people seeing Perplexity as a place of trusted information. Their, you know, announcement came along with the note saying, essentially, we don't want people to be unsure if what they're seeing is a true result or a advertiser sponsored result. And so that idea of trust driving is, I think, really important, and as we all know, I think Anthropic really has placed its bets very much in the trust category, as well as we saw, separate but related topic, their sort of clash with the Pentagon recently, you know, Anthropic really doubling down on trust. And so I think, you know, as we think about this from a CMO perspective, you know, I don't think we're in a world where we can choose either or, right, trust versus reach. I don't think we can sort of carve off any of these LLMs and not pay attention to them or not use them in our marketing. But when I, you know, but when I think about what CMOs should be considering here, I think there's a there's a lot to think about, and I think you really need to consider having strategies for both the trust layer and the reach layer. You know? And that really that really means experimentation with OpenAI's ad platform. It means experimentation with the way Google is extending their long history of advertising and approach to it, but also really leaning into these trust based platforms, right, and treating each one in the way that it is most effective. And, you know, I think that users are starting to really look for, them to behave. So fascinating split starting to take place. Who knows where this will go in another six months, but been an interesting been an interesting couple of couple of months here. Yeah. For sure. I mean, the way I'm thinking about it is almost like there's an organic portion of it, which is, you know, all your LLMs. So you obviously wanna make sure from a brand discovery standpoint, your brand shows up in all LMs. But of particular focus with Claude and with Perplexity. Right? Because they're not doing advertising. And then with Google and ChatGPT where they will be doing advertising, having both an organic and a paid media strategy for those LLMs sounds like. Right? Yeah. Absolutely. And I think, you know, it's like with everything else with AI right now is it's too early to make final bets on anything. You know? And I think understanding how these platforms are handling the two different paths, I think, is really important so that your your brand can show up in ways that feel relevant and interesting. And, again, not ignoring one or the other, but having an approach that that makes sense within the context of your brand. We'll talk in a couple minutes here about another platform that's kind of a bridge between these two strategies. But your point is exactly right. Is, you know, lay some lay some chips on each side of that table as we move forward here. Yeah. I mean, I'd love to get your take personally on do you feel if you see advertising on Gemini and then on ChatGPT, And as long as it is labeled and not intrusive to the search results, does that hurt your sensibilities? Yeah. It's a good question. I think I I feel a couple different ways about it. One, I do think it impacts the experience. Right? We have for the last, you know, however long it's been since ChatGPT launched, two and a half, three years, I guess, have had a pretty unencumbered experience when it comes to these LLMs. We haven't been, you know, faced with advertising and intrusive experiences around that. So I think any shift away from that feels different, and it feels a little bit, maybe not alarming, but certainly unique. So, yes, as I've experienced ads, I've experienced them in Perplexity, I've experienced them in ChatGPT, I've experienced them in Gemini, It does, in my opinion, sort of take away a little bit from that feeling of unencumbered experience in search and conversation. You know, I'm not I'm not naive to the fact that they have to monetize these things, But I guess, you know, I go back to where Sam Altman started maybe two years ago where he thought, or where he said he was hopeful that, you know, there would be new models emerge for supporting LLMs and sort of moving away from the old school Google model. And so it's kind of ironic that OpenAI is is really leaning into the paid model, at a time when I think there are probably other ways that we can that we can do this. Very cool. Yeah. And we'll see how long Perplexity and Anthropic kinda hold off because I feel that when it comes to monetization through advertising, it's just a matter of time. I mean, I feel Anthropic with Claude has more of a intentional strategy about it because they're going after the b to b space, and their margins are bigger. Their bets are stronger in terms of that b to b, so they're not looking at advertising from a b to c standpoint and using that as a model. Perplexity, I'm not entirely sure about what their strategy is, to be honest. So I think that is something that we'll watch. I wouldn't be surprised if they get scooped up by a larger company and, you know, Microsoft or whoever. So let's see how that all shapes out. Yeah. Perplexity has been interesting to watch because I think, you know, you put your finger on it. Right? That I'm not I'm not clear what their strategy is either. Right? I mean, it's yes. We're doing ads. No. We're not doing ads. You know, yes, we're going to launch all these expensive paid models, but we're also going to offer some free services. They're doing a lot in this sort of one click shopping area, partnering with different retail brands related, you know, marketing within their platforms as well. So they feel a little bit like, you know, a thousand a thousand directions all at once, to be honest. But I think Anthropic has been steadfast in their approach to trust. And and I think OpenAI has flip flopped a little bit. And, course, Google has always been about that paid advertising. So a couple of these players are are really focused on clear strategies, and others, I think, are still trying to find their way. So speaking of finding their way, Abby, I wanna talk about lobsters. Are you ready to talk about lobsters today? I know where this is headed, man. Okay. So if if folks are paying attention to, you know, all things AI, you will have seen some red and blue lobsters appearing in your feeds a few weeks ago, and there's been a lot of activity related to those lobsters that Abby's gonna talk through with us here and tell us the very latest about what's going on with our friends with OpenClaw. Yeah. So for those of you who are not kinda queued in on this, so November last year, there's an Austrian developer, Peter Steinberger, who released an open source AI agent, and he called it Clodbot. And then, of course, the folks at Anthropic with their LN Clod got a little heartburn about that, and then he renamed it Moltbot, and then ultimately Openclaw. So they're all the same thing, and Openclaw is what it is right now. And unlike traditional chatbots, Openclaw could take action. You could install it on a Mac Mini. You could execute commands, manage files, interact with applications, automate workflows, all on your behalf. Right? And the growth was explosive. So in under ninety days, it it was almost two hundred thousand GitHub stars becoming one of the kind of fastest growing open source projects in history and gaining hundreds of contributors. And why are we talking about this in the February episode? Because Steinberg ran out that he was joining OpenAI to work on personal agents. Right? So he and Sam Allman got together. OpenCLaw remains an open source project. And the move sort of underscored the fact that all major AI labs were investing serious talent and money into systems that can take actions and not just generate tech. So, it created all kinds of buzz around the industry because it was really performative in terms of the doing part. There were lots of security concerns, and there was also something called MoltBook, which was Reddit style social media platforms where all these open claw vox apparently were getting together. So it got a lot of news. Some of the things were pretty wild that the agents got together and came up with a religion, and they swindled their owners out of money. And how much of that is true and how much of that is not, who's to say? But all this was a little wild and was kind of indicative of if this becomes real in the future, what the future may hold. So exciting, but also a little creepy and scary at the same time. And I was just sort of fascinated by what this may mean for marketing folks. Right? And in a way, for all blue collar roles, if you're in front of a computer and doing things, AI is moving beyond helping you write copy. Like, if you're a marketer, we all use ChatGPT to write copy or ideate or what have you. But systems like OpenClaw show that AI can really take actions inside the tools that you could use. So with OpenClaw, you can connect with it to a WhatsApp or whatever and talk to it and give it instructions, and then it would go execute that on your behalf. So imagine if this all works out fine, that means from from a CMO standpoint, this could mean building campaigns, building reports, updated segments, coordinated tasks across various marketing platforms with a lot less manual effort than what, you know, marketers and campaign folks may be used to. And, you know, when execution speeds up, clarity really becomes the advantage, and that's where, you know, this is highlighting a world where clear goals, defined audiences, stronger creative briefs, the right kind of position is gonna matter more than the actual grunt work of execution, because if inputs are vague, AI will simply execute a vague strategy faster and amplify, which is sort of dangerous. So, in a weird sort of way, AI is going to focus and make execution easier, it's fascinating to see that it would almost focus force the human part out of a marketing strategy where you want to make sure your goals, your tastes, your judgment, and all those things need to be stronger. What's your feeling on on this whole Open Claw news and and noise, I should say? Yeah. Noise is a good word for it because it it's kind of interesting to me that this one bubbled to the top and made so many waves. Mean, I think there were the sort of wild headlines that you described about the social network that they created on their own and, you know, how they were, you know, supposedly communicating amongst themselves and things like that. That to me is probably what immediately made this thing break through. But it it was a little surprising to me that something that is essentially an agentic solution would get this much attention. Right? Because we talk about agents all the time. We talk about the automation of work. We talk about, you know, how much, you know, autonomy are we willing and able to give these agents. You know, I I did a talk at the Marketing AI Institute's AI for Agencies Summit a couple weeks ago on agents and building agents. And there were comments that were running through the thread as I was talking about, you know, isn't isn't what he's talking about really just, you know, workflow automations. Right? And in some cases, that is that is true. Right? I mean, there are things that, you know, are workflow automations. But when you string those things together into something that feels not feels acts agentically and acts on your behalf, that's where that real power and that real unlock be you know, becomes or comes through. But interestingly, like we've been talking about agents and the the sort of guardrails and opportunities for, you know, the better part of a year or more now. So I was just interested to see why this one broke through. I get I'd love to hear your thoughts on why why OpenClaw slash MoltBot slash all the other things that it's been called in its short life. Why, in your opinion, did this one break through when some of the others maybe haven't? I think it was just because of the fact that it was unrestrained. It was so easy easy for the technical folks, maybe not yet for just the common person who's not a developer. But it was super easy where you just install it on a on a Mac mini, and then it has completely unfettered access to all your systems, and it was executing and talking and and some of the demos. I have not installed it myself and played with it, but I have interacted with it with a friend who had it installed. So it's pretty wild where you can just give it instructions that has logic, and the fact that it is not kinda controlled or constrained in any way, I think, is what people really found valuable because you could really kinda, it could send email, it could do things for you. And so I think that leap from just using AI to write things to actually having AI do things in a very tangible, visceral sort of way was, I think, what got everybody's attention. And I think the funny thing is that it was unfettered access to your systems, and, obviously, there were a lot of security issues that came out of that. But to me, a CMO standpoint, that is also a wake up call that when it comes to your AI strategy, like guardrails are no longer optional. Right? If an autonomous system, you know, imagine can access your CRM data, can access your paid media budgets or content libraries, you know, and if it messes up and mistakes can scale quickly, like we know AI is not perfect by any stretch, right? So CMOs need to define those guardrails, define permissions, what is allowed, what is not allowed, visibility into what the system is doing. But then the other part of it is recognition of the fact that if your marketing workflows and processes are confusing and messy, then all AI is going to do is just amplify that. So, you know, clean data becomes a performance multiplier. You want to make sure you have well structured segments, consistent naming, reliable customer data that allow AI to perform intelligently and not just have any sort of agentic system kind of run the streets wild, just connecting to whatever system and doing whatever. Yeah. Yeah. I mean, I think that's the next major frontier, I think, here in marketing, but in AI writ large is the impact that agents are truly gonna have once they're able to really do deliver on the promise that we've been hearing about for a while. So, yeah, agree. Fascinating stuff. So let's switch gears, Matt, and talk a little bit about Reddit. So Reddit is sort of, you know, the single biggest repository of human commentary, as I understand it. And a lot of AI LLM search engines are referencing Reddit in their indexes in order to make sure that they can really examine how, you know, humans are commenting on various topics, the subreddits and what have you. But Reddit is really positioning itself as the antidote to AI driven shopping. Unpack for the listeners what that means, and what are the implications for CMS? Yeah. It's it was a it's a fascinating article. We can share the link. You know, we talked a few minutes ago about the reach versus trust strategies that, you know, OpenAI and Google and Anthropic and Perplexity are are taking on now. And in a lot of ways, Reddit right now is is playing in the yes and strategy category. They are doing both reach and trust and doing it, I think, a in a pretty smart way that bridges the two of those, and we'll see ultimately how that plays out. But at its core, what's happening, you know, everybody knows Reddit is the place where people go and share thoughts and opinions on the most, you know, minute of minutiae. But it's a place where people get together to share, you know, thoughts, opinions, ideas, that real human content that you were talking about a minute ago. And so there's, you know, obviously a natural push, you know, certainly on the side of the the frontier model companies to wanna be engaged with Reddit. And last year or maybe late twenty twenty four, OpenAI struck a deal with Reddit, very lucrative deal, something along the lines of sixty or seventy million dollars a year to use Reddit content there in their responses within their LLMs. And so, you know, at that moment, was like, well, is Reddit selling out here? Are they going to the, you know, to the big LLM company here? And and what does that mean for the community of Reddit and, you know, the sort of culture of Reddit? And I think that's, you know, probably still a little bit of an open question. But they are also very actively, as you said, positioning themselves as the antidote to AI driven shopping in a way that is bringing that human connection and that trust and that reliability very much front and center. And so, you know, I think brands are needing to start looking at Reddit as a really, I think, an interesting an interesting place that they have to have a unique approach to. Right? Because on the one hand, you wanna show up in a legitimate way as a brand within some of these communities. You wanna be seen as contributing to these communities and and making an impact in a positive way, but you also wanna, you know, rely or you also wanna focus on the realities of your business and needing to drive, you know, revenue and things like that. And so I would say that as, you know, as CMOs, you know, start to look at the LLMs, traditional, you know, web searches, and then Reddit as this sort of, you know, center of of all of that is, you know, have an approach to Reddit. Right? Pay attention to what's happening there. See what people are talking about in the areas that matter to you and to your business. Become a real and relevant part of those conversations. You know, don't just use them as straight marketing channels because it will get sniffed out. We've all seen, you know, we've all seen Reddit threads that go south when somebody tries to start selling something. So go there and be a legitimate con you know, connector and contributor, but also pay attention to, you know, how the information that's being created there, the conversations that are being had there are being brought into the LLMs and and to the answers that they provide. So, you know, it's it's a unique place. Reddit has always been a unique place, and it continues to be so in the age of AI. And so I would just say pay some attention to those, you know, those parts of that community that really are most relevant to you and your your business. So with with the NLM sort of indexing Reddit threads or Reddit data Yeah. To train their LLMs on that. So essentially, from a marketing standpoint, what that really means is that you wanna make sure that your brand is present on Reddit. You know? Like you said, not in sort of a selling sort of way, but authentic conversations, making sure that as a brand you are contributing, and definitely the Reddit community will make sure that, you know, if it sounds sales y or kind of promotional in any way, you'll get tagged, flagged, and kind of voted down. So definitely be careful about that. But really, in a lot of ways, this is about brain credibility, making sure that you're kinda facilitating dialogue and making sure that your brain shows up in positive light and which then, in turn, make sure that your the LLMs are indexing that information and then show up in brand discovery when people do AI search. Right? So in a lot of ways, it's sort of. Right, Matt? Yeah. It's a it's you put your finger right on it, Abi. I think, right, it's it's there's a PR component to this. There's a, you know, playing nice in the sandbox and being a good contributing member member of society. It's about making sure that you are visible as a brand. But what's really you know, what's fascinating to me about AI increasingly, what we're seeing is that AI answers and AI itself trust conversations about brands more than they trust brand content. Right? So if someone's talking well about your brand, it's like the old days of, like, you know, when you would when you'd go to OpenTable or, you know, some of these sort of community, you know, community driven apps where you would get legitimate research or responses or reviews about a brand, that makes you as a consumer feel like, okay, this is a legitimate place. I'm willing to give this place a try. And I think we've seen a big shift away from that where people are sort of review bombing these sites, and so those have frankly lost a lot of their trust. But AI at the moment, these LLM results, to your point, really care about true and trusted conversation. They're not interested necessarily in your branded content, talking about yourself, pounding the chest about how great you are. They wanna know that you are seen as a legitimate player in the space. And I think that that line is a really important one for marketers to walk, which is, yes, you wanna promote your brand. Yes. You wanna be visible in the communities that matter to you, but you have to do it in a way that both the community and the AI are gonna see as valuable and beneficial to the overall community. Yeah. It's almost like adding the the value of human opinion as a significant kinda weight to to LLM findability, making sure that, you know, in the list of all the sources that these LLM sort of factor in in in whichever way, making sure that authentic human dialogue is a key part of it. Yeah. Totally. And it goes back to that word that we started with, right, which is trust. Authentic authenticity, as you just said, trust, human connection. In many ways, those things are more important than ever in the age of AI. And so exactly as you said, really, really leaning in to those things that connect people and create community as you do so well with, you know, the marketing AI Pulse community. That kind of thing matters, in my opinion, more than ever. Alright, Abby. On to our next topic, which is a topic that is near and dear to your heart, which is SaaS, something that you are very well known for. Oh, I'm sorry. This says s a a s, not s a s s. Sorry. This is a different kind of SaaS than the ABBYY SaaS. We're gonna talk here about the SaaS pocalypse that has been making the rounds in the world of AI here in the last month or so. Why don't you tell us about the SaaS pocalypse and tell it in a sassy Abbey way, if you don't mind? Thank you, Matt. I would expect nothing less from you, my friend. But but, yeah, I mean, early February, Anthropic introduced advanced automation tools. That's where it all started. Right? It had advanced automation tools and plug ins for its cloud cohort platform, which included, like, legal automation features. And investors saw this as proof that AI could really handle complex tasks normally managed by enterprise software. And that shift in perception or that sort of correlation between what Cloud came out with and what does that mean for enterprise software resulted in a belief that AI might weaken the traditional SaaS market and SaaS subscription model. And the traders reacted quickly, selling off software stocks and erasing almost hundreds of billions of dollars in market value in a single day, right? In mid February, the software sector lost more than like one trillion dollars in value, and you had companies like Salesforce and Adobe into it. HubSpot all dropped as market sentiment sort of moved from more of a predictable growth to a disruption risk. And that decline was driven less by earnings results, but more by a belief that, you know, AI could really reduce that long term demand for these traditional subscription software. And this is early February, but since then, Anthropic has also impacted other industries, right? So it came up with, you know, cybersecurity and the fact that it could scan your code and, you know, identify issues, and all the cyber stocks were getting impacted. And then, I believe this week, it said that, hey, we can help you convert, scan your COBOL code, this code programming language in the 50s. And that resulted in the IBM stock tanking, because, you know, that's a significant part of what IBM does, is take sort of all these legacy systems that exist on that code base, which IBM services in terms of upgrading those. So a lot of what Anthropic's coming up with has impacted this space, but with the SaaS market, it was quite a bit of chaos. Obviously, you can imagine Martech is sort of a key part of, you know, that SaaS space. And it was kind of a wake up call, but not in terms of the stock portion of it. So it's not like a stock story, but more of highlighting a strategy story when it comes to marketing technologies for marketers and CMOs, right? So your martech stack is a strategic decision. It's it's not just an operational one. What that means is that every time you come up for renewal, you should really be kinda questioning, do you really need that tech stack or that specific piece of technology? You know, what are you paying for? Is there a role that AI can play in whatever that technology is trying to achieve? Is there an opportunity to consolidate or enhance leveraging AI? And what are the other new AI native players in the market for that? So the value conversation changes, right? So AI is increasing productivity, and leadership is going to expect that to show up in faster campaigns, in a stronger pipeline, or improved margins. So you need to point to tangible impact of these technologies and not just the fact that it's doing some activity, right? So technology alone is not gonna differentiate you. Everybody has the same access to AI based technology. The real advantage comes from data or brand strength or positioning clarity and the speed of decision making, right? Like that to me is democratizing your ability from a marketing execution standpoint. You know, there used to be a time when you had these big enterprise companies with more sophisticated software that would help you achieve more goals, but now maybe some of the SMBs can have access to the same pieces of technology and and really achieve the same level of scale so that differentiation becomes less about technology. Right? And then vendor selection comes at more of a risk in the world of AI, right? So some platforms are deeply integrating AI into their sort of core platforms, and others are doing sort of AI washing, and it's more at a surface level. So that's where CMOs need to understand which Martech vendors are building the durable advantage into that platform. Right? It sort of brought to light that essentially CMOs and marketing leaders should really be looking at the world of Martech a little bit differently. And, you know, we spoke about this in our last episode as well. Like, you can't now start thinking about, you know, each of these different technologies in sort of an isolated way. You have to think about it in cohesive, orchestrated system, and it wasn't like that concept is new. It was always there even in the world prior to AI. But now with AI, with technologies like MCP, you have a real opportunity to really making sure that various elements of your MRTX stack are working more cohesively in a more of an orchestrated way. Yeah. I totally agree, and it's interesting. You and I have worked, you know, clients who are facing the realities of of the Martech stack every day, right, who are spending significant amounts of their budget, time, and staff energy on managing the tech stack, making decisions about should we re up or not. Right? I think you and I have have each individually and collectively helped people think through some of these decisions, but there probably are CMOs and other marketers who are listening to this right now who are saying, Abby, this is great. Yeah. One trillion dollars worth of value got wiped out, and this is a big picture thing. And I understand that this is a strategy thing, but the fact of the matter is that I'm sitting here today having to make decisions and bets on tools and technologies that are shifting under our feet on a moment's notice. Can you share just some some thoughts on how marketers can make near term decisions at the same time where they have to be thinking big picture longer term fully knowing that the decision they make today may become sort of irrelevant in six to twelve months if that technology gets goes out of date or gets purchased by somebody else. But, like, what are some what are some tangible things that people can do in the minute to help them, you know, proceed in a pretty complex environment? Yeah. Great question. So first off, my first suggestion is don't be don't be the cat chasing the laser pointer. Right? Like, it's very easy. These days, there's new pieces of technology coming out every single day, and it's all great to experiment. I mean, stay tuned, stay informed, and the capabilities and what people are talking about. But some of the more deeply embedded marketing tools need need time and maturity. So don't just jump on the very first thing you see and is really cool. Really take your time to evaluate, hey, why are you wanting a change to begin with, and what is it that you're looking for that your current system does not provide? And then in seeing what's the value that, you know, whatever AI system that has caught your eye is bringing to the table. Is it is it faster? Is it cheaper? Is it saving you time? Is it adding productivity? And all of those are the type of things that people would look at. In the current environment, I'm seeing there's a lot of hype, and it would behoove us to take time to really dive in deeper and not get sort of enamored by all the hype and really, you know, roll back covers, make sure that you're evaluating the martech for the right reasons. So that's number one. Number two is, which is almost like a more foundational piece, which has nothing to do with martech, it's like what are the goals you're after. Right? Like, strategic goals that have been identified in your marketing strategy, making sure that any technology exploration and investment needs to be part of that overarching So at a very foundational level, if you're really looking at, hey, we gotta make sure, you know, we are reevaluating our CRM or our digital asset management system or our marketing automation platform or what have you, that those are things that are gonna move the needle on, you know, your overarching goals. And some of this is foundational infrastructure, but again, making sure that you're taking it on for, you know, for the right reasons. And then finally, I think I would say, you know, governance wise, making sure that how these technologies are really treating your data is a very key piece of it. Right? Is it secure? Are they training their models on your data? You know? Is it compliant with all the SOC two and all these different standards that exist? So making sure that you are not opening yourself to risks that come from these technologies. And I can tell you a lot of these AI Martech vendors are cutting corners in order to, you know, go to the market faster. In some cases, marketing motions and go to market motions almost precede the technology actually being available because they're talking about it, they're drumming up demand because they don't wanna fall behind, and that technology is not fully baked or is in beta or whatever. And those are the things that sort of get flushed out very early on in early demos. I mean, you and me have been part of demos where we've seen some of these things. So yeah. So that's my that's my take on on how marketing leaders should sort of traverse the new wild world of, you know, AI based smart tech. It is a wild world. Wild, wild west. No question. Alright, Matt. Next topic for us was you know, we spoke about Reddit, where there was a lot of, you know, human and authentic content. But on the other end of the spectrum is AI trying to manufacture public sentiment. Right? Yeah. There was some reporting from futurism. Tell us what the reporting was, and what does that mean for marketing folks? Yeah. So this is fascinating, Abby. Right? Because one of the things that's so powerful about AI is its ability to scale and to do things quickly and to do things at a scale and at a pace that I think really, you know, obviously humans can't keep up with. But, you know, there are there are positive sides of that, and there are, of course, potentially downsides to that. And we saw an interesting example of one possible downside in the last week, which is related to some legislation that was being considered in California. The goal of the legislation was to, over time, phase out natural gas appliances with the goal of of reducing the environmental impact over time. Of course, natural gas is not great for the environment, and so they were trying to get these appliances taken out of out of use. But what happened was a technology called Civoclick got involved, and Civoclick describes itself as a disruptive digital advocacy software. And essentially what that is, is it's a way to create a response at scale. They they, in this case, have blurred the line between civic participation and what, you know, what we would think of as sort of manufactured consensus. This is creating a large volume of responses using AI. What's a little bit unclear as you read this article is how much of that response was genuine human response that was amplified by the AI and how much of that response was actually created by the AI. So to give you a sense of it, so CivicClick, what it does is it can fill forms out really quickly and easily for people. It can create emails at scale and essentially get people to respond at volume in ways that make it seem as if there's a groundswell of support when in fact it might just be the technology itself driving scale. And so ultimately, in California here, this legislation was overturned and didn't get passed. And so it really is starting to raise the some of these questions around sort of how do you use AI in positive ways to get people to engage in important topics, but you're not also using it to, you know, persuade people at scale infinitely in ways that could be perceived as negative. And, you know, maybe seen as sort of putting your thumb on the scale, if you will. It also, to me, raises an interesting question around one of the one of the threads that has sort of gone through our entire conversation today is around authenticity, trust. Right? And I think that a technology like this, you know, this is one example, but, you know, the scale that AI can create, really makes something like authenticity that much more scarce and that much more valuable. And so I think there's a real lesson here for the communicators out there, the marketers out there, to find the balance of how they're using these technologies to benefit what they're doing, but not to do so in a way that undercuts their authenticity. So what does that really mean? Like, from a CMO standpoint, you wanna make sure you're authentic and you're not really using AI to, like you alluded to earlier, manufacturer sentiment or manufacturer engagement? Yeah. I think look. I think we all know, and you mentioned it in the last segment, you know, that that we've moved beyond the place where, you know, AI is writing emails for people. What I think we haven't moved past yet is the place where AI can create communication at scale, right, and create massive volume of emails, of social media content, etcetera, etcetera. And so I think marketers and CMOs in particular really need to be cognizant of taking advantage of the scale that's possible and the speed that's possible with AI, but not leaning so fully and so heavily into it that you're undercutting your brand's authenticity and trust that people place in you. Right? So if you're able to say, like, hey, you know what? We could send out, you know, a hundred emails a day at scale to a million different people, man, that's gonna get annoying awfully fast. And I think it's gonna immediately start to undercut the value of your brand. And so I think it's sort of like a double edged sword. Right? It's use these technologies in ways that are beneficial, but don't use them in ways that undercut the way people view your brand and experience your brand. So don't spin up a hundred Reddit agents posting on behalf of your brand to make you sound to make your brand sound positive and because that feedback is not authentic, sounds like. Yeah. It's, you know, it's interesting. I was I was looking the other day at so I'm a New England Patriots fan. Unfortunately, my Patriots didn't perform as well in the Super Bowl as I was hoping they would. But in advance of the Super Bowl, I was looking online to buy a Patriot sweatshirt. And I went to a website and looked at the and I looked at the the comments and the reviews that were there, and it had, you know, I wanna say tens of thousands of positive reviews. And I thought, okay. This is cool. I'm gonna read through some of these reviews and see what they're saying about the quality of the sweatshirt and, you know, the thickness of it and the sizing and all the stuff we look for. And as I read through these comments, it became very clear that each and every one of these reviews had been created by AI, because they all basically followed the same pattern of describing the quality of the sweatshirt, the speed with which it was delivered, and the value that they felt like they got from the purchase. And so as you read through these and you're reading literally dozens of these after one another, it became clear to me that this was something that had been produced at scale by AI. And so when I look at what the CivicLick example, there's a lot of there's a lot of similarities there, which is, yes, you can do this at scale, it's gonna become very clear to your audience very quickly that it's not real. And the the the impact for me personally in that that purchasing moment was I'm actually not gonna purchase a sweatshirt from this website because if they've got all these AI driven BS reviews, that makes me question their brand at its core, it's truly a good value and good quality or not. And so I think there's something to be said here about scaling and using that power in in tangible ways, but using it also cautiously because people will will sniff it out pretty quickly. So what does that mean for CMOs? Governance, transparency within their within their brand or for their brand within their team? Yeah. I would say I would say governance. I would say caution. I would say paying attention to the the the human qualities of your brand, and and having the the things that matter to you as a brand be articulated in how you market to the world. Right. So, again, if we talk back to to Reddit example, it's taking part in the conversations in legitimate ways and not just using Reddit as a platform for your marketing and for your branded content, but as a way for authentic connection and communication. And I would say in the same in the same vein, if you're going to use AI to help you amplify your brand or a message or a campaign, do it sparingly and do it strategically in ways that don't undercut the way people see you. Makes sense. Alright, Abby. We are moving on to our final our final topic for this month. Interesting story coming out of New York State late last year. The the state of New York passed a law requiring, clear disclosure when an ad uses AI, And I think this is probably something that most of the marketers out there wrestle with every day. I know I hear about this from my clients, and I know, in fact, you and I have worked on a project where we helped a client make very clear guidance and guardrails around the use of AI in their advertising and marketing efforts. So tell us a little bit more about what's going on in New York and what impact this is going to have for marketers everywhere. Yeah, absolutely. And so like you said, in late December, New York passed law requiring clear disclosure when an ad uses AI generated human like spokesperson, right? So, avatars, digital influencers, synthetic presenters, basically, that look like real people and could reasonably be mistaken for a real human being. And the law takes effect in June ninth, which is why we are talking about it in our February episode right now, making sure that there's a lot of work going on for brands to prepare for this, before the logs sort of kicks in, which means that ads shown in New York to New York audiences should really must include a, quote, unquote, conspicuous disclosure if a synthetic human is used in the actual advertising. And, honestly, New York isn't the only state. You know, California, Massachusetts, Georgia all have or it's in the works with additional AI transparency and disclosure proposals emerging. And there's more states that are thinking and planning that that sort of strategy of making sure that you are letting everybody know when people in your ads are AI generated. So what that means for CMOs is accountability has really shifted to to marketing departments. So AI is not just a performance lever. It is now brand and compliance exposure in the market, meaning CMOs need to own whether AI driven content claims targeting synthetic personas, people could mislead customers and erode trust. Here we go. The trust word pops up again, map. Right? You know, also I feel that most companies have some sort of AI governance and policies handed down to marketing departments. They're written for internal use, and they're written from a legal and compliance and IT infrastructure standpoint. I feel that marketers really need to think about their own governance, not just an enterprise handout that they inherit. I'm not saying that's not important, but there needs to be one more layer of marketing defined governance, which would apply to marketing employees, to agencies, to contractors, right, where you have marketing specific rules for what is allowed, what is high risk, who approves what, what do you need to disclose, and those sort of things. And it's not just that, there's also a lot of cases going on at the federal level where FTC, the scrutiny of deceptive practices, where you want to make sure that, as a marketing department, you are disclosing very clearly. Now, the New York law doesn't, it says conspicuous disclosure. What that means is TBD and how you interpret that, But regardless, this is something you can't really, you know, ignore. And me, personally, I think that's a good thing. I think it's good to let people know when you're using AI, when it comes to video content where you are clearly indicating that this is not a real person. So I think that's a good thing. And I've seen in a lot of ads where they sort of use the generated content much along the lines of the previous story we were talking about where especially if you have AI generated personas doubling up as, quote unquote, users of your software endorsing your product and the whole premise is all synthetic. Again, that sort of definitely erodes trust in in your brand and will sort of question how people perceive your brand. What's your what's your take on this, Matt? Yeah. I think there's a there's a lot here, probably a topic for another day, but I'm I'm fascinated by the separation of or the emerging challenge between state and federal guidance and guardrails and laws when it comes to AI. There's a real power struggle that's going on between the state and federal governments right now in that regard that will play itself out over time. But the thing that immediately comes to mind for me as we think about, you know, how CMOs and other marketers can think about this is, again, something you and I have counseled our clients on, which is it's best to start somewhere when it comes to governance and guardrails and rules than to wait for this situation to settle itself out, and then, you know, sort of create a a final list of of perfect rules and regulations. All of this stuff is moving too fast. It's changing too quickly. And, you know, I always one of my favorite phrases in the age of AI is don't let the perfect be the enemy of the good. And I would say that what I like about what New York is doing, what California has been doing, is that they are starting somewhere. They are putting a stake in the ground. Yes. That stake is gonna move. Yes. Things are gonna evolve over time, but they're starting somewhere. And I would really just say to marketers out there that it's so important to start somewhere in your own creation of guardrails, guidelines, and governance because, you know, to your point, the responsibility is going to lie with the CMO if there's an issue. And so, your policy might need to change three months after you release it, but at least you have something that you can rely on that your teams can rely on and that your customers can rely on. And so, you know, that to me is an area that I feel very passionately about, which is start somewhere, evolve as you need to. And I know this is something on your side. You obviously agree with and counsel your your clients in as well. But any immediate tips you would give people to, you know, to not letting the per the perfect be the enemy of good? How how can people get started with something if they haven't yet? Yeah. I mean, when it comes to marketing specific governance, I say, you know, people get really overwhelmed by that, but I think I have a more simplistic approach to that where I tell people to think about it in terms of sort of three pillars, like input, output, disclosure. So input is be sensible about what kind of stuff you're putting into AI. Don't download, you know, your your Salesforce data in a CSV and upload it into ChatGPT and start asking questions. Right? That is not something that you should be doing. So be very clear and have a documented policy on on rules for input, what you're putting into AI. Some of that is gonna come from your IT and your compliance people, but also have a have a way to think about this in terms of marketing. Output. Output would be making sure that you are establishing governance and rules around what is a human obligation of what needs to be reviewed when things come out of AI. We know AI hallucinates, so fact check it. Make sure it is compliant with your brand voice and your brand values. If you work in regulated industries, make sure is compliant with all those rules and those sort of things. Right? So making sure that you're not just taking output AI output at face value. And the third thing, which is sort of the the core crux of the New York AI disclosure law, is disclosure, is where you're telling people where AI was used, how it was used, and you're telling people about it in the correct way. Like, there is on a scale of a zero to a hundred, you can completely skip disclosure, which I would not recommend. But then you can be very subtle about it, where you can have it, you know, in fine print on a video or whatever. But we are seeing more and more where as the law starts to catch up with the pace of AI innovation, the governments are gonna start really mandating how these disclosures are done so that people can start really trusting that AI content. And, obviously, this episode is all about marketing, but think about the value of that in society at large. Like, what if you start seeing commercials about politicians and all that kind of stuff, right, where you wanna make sure there's elections coming up, you wanna make sure you can trust some of this content and make sure that, you know, people don't assume things based on whatever AI based storytelling that is completely synthetic, aka fake, that, you know, that is processed by people. So Yep. Yeah. Great advice. Good stuff. Great way to wrap it up. Yeah, absolutely. That is a wrap, Matt. Thank you very much. Enjoyed this episode and appreciate your time. Hope you feel a little bit better. Thank you. Yep. We're getting there. I'll be all good. But, yeah, looking forward to catching back up in March and seeing what's happening in the world of AI. Awesome. To our listeners, thank you for tuning in. Be sure to follow Matt and me on LinkedIn. If you wanna check out more episodes of the podcast, you can go to marketing AI spark cast dot com. If you wanna follow the Marketing AI Pulse community, you can go to marketing AI pulse dot com. Till next time, thank you very much. Thanks for listening. Big thanks to SparkNovus and the Marketing AI Pulse community. You can join the community at Marketing AI Pulse dot com.