WFM Unfiltered

What does the Law say about AI | John Walter

• John Walter • Season 1 • Episode 33

Share episode

Message the show!

What does the law say about AI? And more importantly, what should we be asking? In this episode of WFM Unfiltered, Irina sits down with John Walter — lawyer, tech innovator, and founder of ProxyLink — to dissect the growing tension between artificial intelligence and legal boundaries in the contact center space.

From transitioning out of the legal field and diving into customer support, John’s path has been anything but typical. His frustration with how AI is being rolled out (and oversold) in CX platforms led him to create ProxyLink: a solution bridging secure communication between AI agents and humans. But this episode isn’t a product pitch — it’s a provocative conversation about rights, risks, and the very real legal traps many businesses are sleepwalking into.

You’ll hear surprising insights into how bots could affect liability during emergency calls, why consumers need advocates in the AI space, and how human-first design might be the only path forward. Expect practical advice on regulation, actionable tips for CX leaders navigating automation, and some spicy truths about “AI” apps that aren’t really AI at all.

If you're a decision-maker in customer service, CX, legal, or tech — this is your wake-up call. Get ready for a conversation that challenges buzzwords and exposes the legal blindspots no one is talking about.

Subscribe to never miss an episode: https://www.youtube.com/@wfmunfiltered?sub_confirmation=1

Show Links:

RightWFM website: www.rightwfm.com
Email: Irina@rightwfm.com
Podcast email: WFMUnfiltered@gmail.com
Podcast Directory listings: www.wfmunfiltered.buzzsprout.com/share
YouTube Channel: www.youtube.com/@WFMUnfiltered?sub_confirmation=1
Guest LinkedIn: www.linkedin.com/in/jowalter
CCAIA: www.ccaia.org
ProxyLink: www.proxylink.co

Thanks for listening.

If you'd like to contact me about the show, you can email me HERE.

If you have questions about working with me on WFM projects and Consulting, you can find The RightWFM website HERE.

Please remember to subscribe and leave a review of you've enjoyed the show!

Irina:

Hey everyone, and happy. What day is it today, John? Happy Wednesday. We're pretending that it's Tuesday today, so happy Wednesday everyone, and welcome to another episode of WFM Unfiltered. I'm your host and I have someone special today in store for us. But before I introduce you to my guest. As a quick reminder, please hit that subscribe button is really helping the show. And if you're in need of any WFM assistance, you can check RightWFM.com or reach out directly to me. And with that, enough of the business, let's move into the topic of today and to the guest of today. Hey John, how are you doing?

John Walter:

Super grateful to be here and to be talking on your podcast. It's very proud of you for what you've started here. I think it's awesome.

Irina:

I, I just wanna make a quick note that I think John. Yours was the first podcast that I've ever been a part of, and I was absolutely terrified and horrified, and you did such a great job of just calming me down and put my, my stress behind me. So thank you so much for actually starting the, the journey for me. I'm the one who is so grateful for you.

John Walter:

That's really cool. And, and I ended up discontinuing my podcast when I started the contact Center AI Association, just because I got too busy, got too busy to juggle it all. It's a lot, it's a lot of work. And so I just love, I love, love, love. It's, it's, it's fun now to be on the receiving end of, of of, of your podcast. And I, I, yeah, it's a big fan of, of what you're doing.

Irina:

Thank you so much and a bit of a disclaimer, I'm not so kind as you are. So I'm used to throw curve balls at my guests because I wanna spark a very good conversation. But I don't wanna jump into the topic before I give you the opportunity to share a little bit about your background.

John Walter:

Yeah, thank you. So my name's John Walter. I practiced law for almost a decade before entering into the customer support industry. One of my clients in the legal practice was involved in customer support outsourcing, and that's originally what got me involved in the industry About two months after I started working full time in, in the customer support industry, I saw the arrival of chat GPT, and we just saw that as being a very, monumental. Technology as that would be very influential to the industry. And so that just led me to start diving into that and having a lot of conversations that that's what inspired my initial podcast. And, but all of those conversations led to the creation of the Contact Center AI Association, which is a 501C3 nonprofit, has a charitable purpose of education. And all we do is bring together customer support and CX leaders to come. To close door settings where they can honestly talk about the issues they're facing, what's working, what's not working in an environment that's safe, and, and where nothing's broadcast and nothing's recorded. that's where we distill a lot of insights about what's, what's, what are the real pain points that are facing the industry and, and how companies are going about solving it in their internal operations. And then our goal is to clean up. upon those insights and distribute them as broad as we can throughout the industry for free. That's our, that's our mission. And so that's been going on for a little over a year. We have chapters in Atlanta, New York City, and then we had our first chapter meeting in. San Jose, California, just south of San Francisco a week or two ago, I think that was last week. And then we're gonna, in Q4, we're gonna have our, we plan to open a chapter in Austin, Texas. And then we're trying to get better about having a cadence of LinkedIn live events, more public facing content. And, and so that's something that we're, we're, we're striving towards. that's a lot of my time is helping with that association. We have, we have probably about. 20 to 25 prominent customer support leaders at various organizations. Not, not on the vendor side, more on the, the, the, I call'em in-house. I call'em in-house leaders, just'cause I come from the laws. You have like in-house counsel, you have outside counsel, so I call'em in-house. I call'em the, in the in-house leaders that vol, they volunteer in variety of capacities. And so any, if anybody here is listening. There's ways to get involved, whether it's you can, we have an education committee that meets once a month for 30 minutes, about a dozen CX leaders hop on a Zoom call and identify, Hey, here's the, you, you won't believe what happened to us this month. And, and, and, we need to develop some educational content around this. And then everyone kind of talks about, it's oh, hey, here's some, here's someone I know who has relevant experience, we can bring in as, as a thought leader on this topic. So that's how that works. And then we have a steering committee. It's also, it's a little bit smaller right now, but it, it, it could have some room to grow a little bit more where it, it, it just guides the, the structure and the mission of the association. So they help us make decisions on issues. Okay, if, if, if an, if, if somebody in the community for a product recommendation. Slack channel, like how do we handle that? Like we, we wanna make sure that no one's trying to monetize that opportunity and within this community and, and making sure it's purely education and, and it's, and I'm sorry to, I I, it, been baked into my mind. It's, it's, it's almost like when you're. When you're trying to get testimony from a witness, you you want to, to eliminate bias. And to the extent there is bias, you want to call it out and try to remove it from the equation. And so it's a similar thing happening in that context. Um, then finally, I know it's a bit of a tirade and this is why I might have dark circles under the eyes'cause I do a lot.'cause in addition to the, the association got this, year ago, I just came across this issue in the industry that I found very interesting, which is inbound calls, chats, and emails from AI that's hitting the call center. how do companies make sure that their operations are secure and efficient with those types of And so started ProxyLink as a company that all we do is authenticate. We help companies authenticate and bound AI communications.

Irina:

I am so impressed with everything that you're doing, and most of all, I don't think I have ever heard about a lawyer that doing something with AI is just, it is, I don't know, a very strange combination, but one that it's definitely needed in the market because that's one of the things that we spoke a little bit about it. I'm highly annoyed when the topics about AI comes up because I'm of the belief that a lot of people slash companies are coming up with the, the buzzword of AI and are selling you air and complete bs. And I always freaking hate when you cannot give me specific examples. So another disclaimer. I love AI, I love automation, and I. Hate people that cannot give me examples and are just trying to gain more popularity for their stupid company or business that does nothing for you. Luckily, John, you have some stuff to tell me about it,

John Walter:

Yeah. Yeah. The I, I can share some specific examples, but something that's funny on that topic. When I was in San Francisco last week, I was at an event and it was somebody I was talking with who was. He was at a large, larger tech company that, that builds and, and sells AI technology into the enterprise. And, and he was telling me how their marketing director came over to him and was saying, okay, we need to start saying that we have agentic AI. That we, that we have to have AI agents. And he was like, but we don't have AI agents. And, and the marketing director's no, but we need to say this. We need, and I think a lot of companies feel that pressure each quarter or each month, however frequently it is, as there is this evolution. 2025 is definitely the year of AI agents. Here's a real, real world use case of AI agents that it's a bit different. You know what, actually, you know what, you might might not even call it an agent. You know what, what, here's what I'll say. It's an agentic experience for a consumer. And so when, when you think of an agent. An agent is something that can take something off your plate and handle a real world task. At least that's the way I define it. Maybe I'm wrong. It's, it's somebody, so even the word agent, it comes down to the, the, the concept of the agency law. You have a principle and an agent. The principle directs the agent to take action, real world action on the, the principal's behalf. And what, what's happening. And the consumer space is that despite great effort by many companies to deliver high quality customer experiences, many cases customers still find customer support to be frustrating. And that's a very kind way of ex explaining the industry. And what's happened is there's been this proliferation of consumer rights apps that allow consumers to delegate specific tasks that consumers find frustrating. I. One of those, for example, is subscription cancellation. are services like Rocket Money, bill Shark, and now a proliferation of others like Experian. I hear that McAfee, I, I McAfee, the last time I heard about McAfee, they were doing these like, antivirus, like computer, malware, virus. But now apparently they're canceling subscriptions now too. They, they make it easy for a consumer to delegate this task of subscription cancellation and part of the way they do this. Is by introducing the opportunity to cancel at a time that's most convenient to the consumer. So it might be when they are using their budgeting app and they're identifying, okay, I'm falling short on meeting this goal. I need to cut some expenses. And the app, right? Whether it's a banking app or whether it's a separate FinTech app, you can, in many cases now, consumers can cancel subscriptions just by pushing a button. And so I won't go so far as to say this is maybe, yeah. I wouldn't say that this is like a agentic AI from like a textbook definition that people might to it. Actually, I'll say it's definitely not what, what, what's happening is consumers are having this experience of being able to delegate a task to an agent. So take out the AI part. Take out the AI part. They're delegating this task to rocket money. And rocket money an agent, not an AI agent, is because the terms of service between the consumer. This is, this would be a good title for this podcast episode, like Real Life Agents, real World Agents, and you can take a the AI out of it, like real In the

Irina:

I'll

John Walter:

let

Irina:

you name the episode

John Walter:

Yes.

Irina:

I'm gonna give you the right, you're the one and only person that will ever name an episode on the show!.

John Walter:

And and so what's, so what's happening is, so the terms of service between the consumer and rocket Money, for example, constitutes an agency agreement, which is so fascinating. And so the consumer, when they download the app, the consumer becomes a principle and Rocket money becomes their agent. the consumer has certain legal rights and privileges as a consumer under federal and state law, of course, but then also under the contract that the consumer had previously entered into with the business that they're trying to cancel that subscription with. And when they enter into that agency agreement by signing that terms of service or clicking the terms of service that they agree to when they download the app for Rocket Money or any of the others that do this service. The, the legal rights and privileges of the consumer are then bestowed rocket money. Isn't that fascinating? And then Rocket Money has the ability to call this company on behalf of the consumer or to, or to transact with this company on behalf of the consumer. And now what's starting to happen is that all of these companies. That do these tasks for consumers are use, they're starting to use AI bots to accomplish these transactions, whether it's over the phone, whether it's through email, whether it's through chat, whether they're logging into the website and doing it on the consumer's behalf, and, and so that's. That's where I, that's, I've been fascinated with this topic for almost a, I've been, I've been working full time other than the, the, my duties with the association, the, the Contact Center AI Association, which just volunteer non compensated duties. My full-time vocation has been trying to solve that problem of, okay, how do, how does a call center healthy boundaries AI bots that try to call them? Because, because, let's be honest, companies should be prioritizing human support for humans. Not, not for, not for bots. Bots are infinitely patient. They are. They are. And, and, and, and they, and they also need to go through enhanced, heightened scrutiny for security purposes. And, and yeah. And

Irina:

i, I, I wanna take you back because my mind starts to go in so many directions that I'm really fascinated about. But let, let's take a step back. Currently the hype, especially in contact centers, is, okay, let's cut from the costs of having agents do stuff that bots or AI agents or whatever AI can do for you, which is great because I'm a big supporter of if something can be automated, automated. Why does somebody have to lose their time? Typing something manually that can be done in seconds and save you money, save efficiency and use your people to do something more important. Analyze the business. Think of strategy, whatever it is, but don't do the, the stupid routine tasks. The other idea is, okay, if something can save time for the customer, that's also better for the business. Now the problem is that in. Currently the state of using bots in a lot of contact center is that I, as a customer, am getting frustrated because I cannot get to a human to explain exactly what I want. My use case is not in the menu, and I'm spending 15 minutes more trying to get the bots to tell me what do I need to do with my case. And then I'm starting to get extremely frustrated with the company. So how, how do we work around that first scenario?

John Walter:

Yeah, it's, it is very, it's very, very, very interesting. You're bringing up v various different points when it comes to and, and I, I want to clarify the question a bit because we could go, go to, go down different paths.

Irina:

Mm-hmm.

John Walter:

O one thing I will say is there, there are certain ticket types Like you said, it makes sense to be automated in all cases on the company side. And that's, and and I'm not, so far what I've, my observation has been in the customer support space is that most, most contact centers are struggling to keep pace with the volume of support that comes in and. Yes. And AI is helping them actually breathe for the first time and helping them get their head above water and, and, and actually, adequately manage the volume that's coming in the and, and but on the topic of, on the topic of consumer preferences,

Irina:

Mm-hmm.

John Walter:

There is, there is. There is a, a customer experience component to this when it comes to consumers delegating tasks to third party. This is a premium service that consumers are paying for. Bill Sharkk charges$5 per cancellation.$5, and it doesn't cost that, it doesn't take that much time to click around on a website to cancel something and do not pay is another one of these services. They charge$18 a month. Rocket money, I think charges around like 10 to$12 a month. And, and so there is a lot of consumer demand for a more frictionless customer experience where they don't want to handle it, they don't want to handle it synchronously. It's, we're forcing customers to sit there with us and hold our hand the whole time The company is helping them resolve their problem, when in reality it's not urgent. The consumer just wants it being taken care of. They can, it can be, even if it's something that's relatively complex, it can be handled asynchronously and kinda at the pace of email. But email is somewhat of an awkward customer support channel because it has a reputation for being slow it, and it's, and it's a little bit more laborious. You have to sit there and think and type it out and thoroughly explain things. You send it, you wait 24 hours and then something comes back. But imagine a circumstance where a consumer have a fully vetted conversation with their own AI assistant fully explaining the concepts, and this AI assistant knows all of the facts that. Needs to be retrieved from the consumer in order to efficiently resolve that ticket. so large language models are fabulous at this fact gathering. So you can tell them, it's Hey, I need you to gather these 10 pieces of information and then, and don't stop asking questions until you, you, you get all, all 10 pieces. And, and so imagine the consumer's AI, so on the, the consumer's own device.

Irina:

Mm-hmm.

John Walter:

That, that, that the consumer loves and trusts and perhaps used for a variety of purposes, not just for resolving problems. Maybe they talk to this AI about relationship problems or, or their, or their, their, their fitness goals, or who knows what, and all of a sudden they're like, oh yeah, I need to, I need to do this task. And, and, and I think we're heading towards a future where the, the AI bots are gonna be able to say, oh, yeah to answer this question, I'm just gonna need to retrieve from this consumer, these qualifying information pieces of information, then we'll be able to submit this request to the contact center in a way that is highly efficient for the contact center to resolve. So it is received. So when, when one of these pro I call'em proxy requests, when one of these proxy requests hits the contact center, the the authentication's already resolved. The, all of the information, the call, call center agent needs to complete The task is there on the, on the screen, and, and then, and then there's limited actions that the call center agent can take based on the type of ticket. And so it's, it, it is, it's extremely efficient for humans to manually resolve.

Irina:

I, I, I love that part about the efficiency, and as you mentioned for me. I think the, the, the problem becomes when we're starting to say that bots will completely eliminate humans because maybe, maybe in the future that will happen. We're not at this stage and everyone that's saying that this is gonna be the case in five months, one year.

John Walter:

No.

Irina:

I don't believe it because there's, the other thing that I wanted to discuss with you is the legal part. What do we do when technology fails? We're, we're fighting that terms of agreement and then somebody hugs the system, the system is down, or there is a glitch and something happens.'cause we're working with technology. We know that this happens all the time and.

John Walter:

Yeah. And, and, and going back to this that I was, I was, I was just describing and, and we can explore it in different scenarios also because I think there's a lot of, there's a lot of legal implications all throughout the industry. And this is a fast, so the contact center AI association, one of the pillars of our work. Is what we call the the quarterly legal update. I got a lot of heat for this. When I was first starting the association, I was like, Hey guys, we need a quarterly legal update. And everyone's dude, you're saying that'cause you're a lawyer. And I was like, no, no, Doug, trust me. And they're like, no one's gonna be interested. We need it. It is insane. The, the, the, the, the volume and the stakes for the, the litigation in the United States surrounding the use of AI and customer support operations, the big brands using technology from big vendors you think people would think these things through. And then what's happening is you have these creative plaintiff's lawyers pulling up. State laws from 20, 30 years ago that deal with wiretapping or deal with biometric information and, and they're, and they, and they're saying, oh, maybe this will apply. And I was, I can go ahead and give like a tip here, like a little pro tip for the, the way these plaintiffs are finding their is they are at the patents of the vendors. for the biometrics cases. This is a, this is like low hanging fruit for the biometrics cases, which are mostly coming out of Illinois. They are in, in Illinois. It is illegal to use somebody's to capture, to capture someone's biometric identifiers, which includes a voice print. their prior written consent. With their written consent. So getting their consent over the phone that you're recording it and using AI, insufficient written consent. And, and, and what's happening is these plaintiff's lawyers can look at patents of AI vendors in the call center space and, and, and their patents. All of these, all of these, are trying to sound like they're doing all these like really cool, innovative, and unique things. And so they're putting in there that they're using biometrics and voice prints in their patents all a plaintiff lawyer needs to do is just say, look, it's in the patent. And, and maybe at trial you can prove, okay, no, we're not actually capturing voice prints in the way that, would infringe in this law. it's not gonna go all these, none of these cases are going to trial. They're all settling. They're all settling because the stakes are so high. I believe my recollection serves me correctly. I think in Illinois it's$2,000 per violation for every, every time it's, it's done negligently. And it's$5,000 per violation for every time. For every time that you knowingly.

Irina:

Mm-hmm.

John Walter:

Knowingly capture someone's voice print. And, and so none of the so to do the math, it's, it's take every call from the state of Illinois and multiply it by 5,000. And so none of these cases are going to trial. They're settling for, for probably tens of millions of dollars. so anyways, go back to the way they're getting their targets. And I know, I'm going on a tie right here. This is fascinating. they, I identify the, the vendors that have voice prints or biometrics state identified in their patents

Irina:

Mm-hmm.

John Walter:

In particular, that are handling voices, inbound voice like technology for inbound voice calls. And then they find the companies, the brands. are publicly associated with that vendor. So a lot of vendors will have client testimonials on their website or they'll have, or they'll, they'll have the, they'll have a conference, they'll put on a conference, an annual conference, and they'll invite the brands to come and speak at their conference about how they're using their technology. even had one case last year, it's actually honestly sad. A prominent retailer in the US was sued for using a prominent. technology and customer support, and the CX manager was by name in the class action lawsuit because, because they said that because they said, this person, the CX manager for this entity, admitted to using this technology. And as, as the evidence for that admission quotes, I'll say anyone listening, that admission in quotes. Was there testimonial on the vendor's website?

Irina:

Yay. Wow. Wow. And one thing that they're gonna say is that, i, I'm currently dealing for the past, I don't know how many years with deploying WFM tools in a lot of countries. And I would say that the laws in Europe are so strict and so tight that even traditional WFM tool that has some slight deviations from the law. No, no, not happening.

John Walter:

Yeah.

Irina:

just imagine if we're talking about AI, because we currently don't have. Actually many laws allowing for AI. We're just basically trying to wing it and see where we can get away with it in what country and which not. So I'm always fascinated by companies that are speaking directly to the efficiency about AI when you don't even know whether you can deploy it or how can you deploy it and use it.

John Walter:

That's a, that's a great point. That's a great point. And, and and I don't like legal seminars. And I, and I was attending one recently, and I was just, my, my, my only reflection was like. lawyers bring no value to society, like no pro productive value to society. They're not like, they're not moving except for the ones that are doing like important, like they're saving communities from like poisonous water and, and like actually, like there are some lawyers that are like really at least when it like really fighting the good fight. But when it comes to a lot of profession, it's just kind of like. Making sure everyone's following these rules that sometimes are not actually preventing any actual harm. And so I don't think a lot in a lot of these cases that there's actual consumer harm. And I don't think any, I don't think any of these AI vendors or companies using their tool have malicious intent. think they genuinely are trying to serve their customers. And you have greedy plaintiff's lawyers who are trying to be opportunistic. All of that said, I'm an advocate for, for adopting this technology and, and running forward with it, any risk of litigation and being smart about it. Being smart about it. Just don't, don't publicly identify, don't publicly identify on your website that you're abusing it unless you're a hundred percent positive. That is a very low risk technology proxy link. like Proxylink!

Irina:

Because I also have one problem with this. I also don't think that there is a company that's gonna be malicious in the way that they're using technology process or whatever. My problem is that we're speaking on behalf of the customer, and I'm always saying, I am a customer as well. When I get a call and I get the stupid repetitive message that I know, it's both. I get so frustrated or when I cannot get to a human when I want, when as a customer, I pay you. I want to go to a human, and there, there is no possibility for that. So I guess my problem is. What if from a psychological perspective, which I think even sum up one was talking about it. Sometimes we just wanna speak to humans. That's it. I don't care how passionate or adequate or good your body is, I just want to know that there is a human. What do we do then? I.

John Walter:

Yeah, yeah, yeah. Escalation to a human is, is, is as fast as possible, is super important when, when a consumer desires that. But I, I have a question for you. I'm, I'm curious on, on your perspective, on, you, you, you hit on the frustration that consumers face when they hit a bot, just because it's an inconvenience to their day. Do you think that there's additional frustration on how that. Call may be used. Let's say that call is used as training data for an AI model. No. No personal information is used. No, no. Nothing private or proprietary. Nothing that can be linked back to the consumer. But do you, do you think that raises like almost an ickiness factor with an average consumer?

Irina:

So I've been spending a lot of time to reflect for myself why I am so annoyed with bots, and maybe it's the level that bots are at the moment. Maybe because I know how a lot of companies are using data and how is it stored, and sometimes it's also because from legal perspective. I want to know that I can look for my rights and I want to know that there is gonna be someone responsible that that cannot hide behind technology and some kind of agreements or user agreements. So I think first of all, for me, interaction with humans, I don't know it, it's very important. I will talk to him for a bot or for yeah, assistant if it's something that I truly do not care about. It's book me an appointment at a hair salon, or whatever it is. Like I can use very, very fast. But if I decide to approach you or your business and company and I'm deciding that it's okay for me to spend the time with you, then I wanna know that you're going to be respectful enough to face me.

John Walter:

Is such a good point. That is such a good point. Yeah. It's funny at the most recent. meeting the one in, in San Jose last week, of the call center leaders just asked the group, ha, has anybody here ever had a positive experience with a chatbot like ever, in any circumstance? And, and nobody, nobody could say anything positive. And, and, and, and several prominent companies were represented at the table. The, the, but one caveat though, which is very interesting, two people brought this up, is that United Airlines, increased the quality of their proactive notifications to consumers. So I, I was flying on United out there because I'm, I'm in the Southeast United States.

Irina:

Mm-hmm.

John Walter:

And, and what happened was when I boarded, when I, when I, when my first flight landed in, I think Denver, right? When I take my phone off airplane mode, I have a text message from United waiting for me saying, Hey, you're arriving at this gate. Your next flight is taking off at this other gate, and this is how long it will take you to walk there. I was like, man, that's good. That was a good customer experience. It uses automation, it uses AI. I'm sure in some, maybe, actually, maybe no AI is involved in that. Who knows? But it's a, it's an automated customer experience that was absolutely fantastic. Similar, someone and someone else had the same thing to say about United. So I, I applaud United for that, what they've done there. Secondly, there, there was somebody at the table who said that they got a text message from their barber. And said, Hey how's it going? I, I, I, I, I noticed it's been about like a, I don't know how long they've been, a month and a half since you've come in. Would you be interested in stopping in for a haircut? We have, have, we have some availability this week. it was actually, he was like, oh yeah, sure. He is all through SMS. He's sure, I'd love to. And he scheduled a time and he says it wasn't until he's like halfway through the conversation. That he realized he was talking with a bot and, but he said he really appreciated it. He really appreciated that because he ended up, he did need a haircut because it, because the bot was reaching out to him at the appropriate time. Right.

Irina:

Mm-hmm.

John Walter:

You know, they, they, they had the algorithm in place to know how frequently this person gets haircuts and, and it was it was, it was convenient to.

Irina:

Mm-hmm.

John Walter:

So, so, yeah. I think there's a, there's a time for it, but it's, but I agree with you a hundred percent. The way it's been approached over the past decade has been awful. It's been absolutely awful.

Irina:

I think we're just not considering human psychology and actually the whole time that we're speaking, I'm dying to ask you a question that I'm thinking about for the last good, probably eight months, and we're speaking about the United States right now, but I guess this is a question that's valid in every single country. You know that there are a lot of cases where you're calling nine one one, and the call is not handled in the correct way by the dispatcher, right? And they have to type in manually all the details that they're hearing on the phone, and they need to contact the, the police or an ambulance or whatever, and they need to stay with you on the phone. I'm wondering, in those high stress emergency situations, how many people will opt in to talk to a, to a bot versus a human? And is it

John Walter:

would, uh uh, actually it sounds crazy. I actually had to call 9 1 1 this morning. Yeah. I called, I had to call 9 1 1 this morning and, and I got a human. If I got a bot, I would've, I would've flipped it. I would've felt like society has failed

Irina:

mm-hmm.

John Walter:

me, Dramatically failed me this morning, and, and so it's, it's, yeah. Yeah. That there, there, there is a lot of circumstances where that's absolutely. Unacceptable. But flip side of this is, like you said, there's a lot of disclosures that need to be made by the, by the representative. There's a lot of typing that needs to be done by the representative. That should a hundred percent all be done by AI. Yeah, you should have AI, know, keeping a transcript of the conversation and making like a checkbox, checking the boxes, making sure all the disclosures are, are made if they're not made, providing visual reminders to the agent to make those disclosures. And, and they should be just transcribing and summarizing all of these calls. Call center agents have a very tough job. And, and, it's interesting. I think it can depend on, I've heard some people say that when you eliminate those types of tasks, you're eliminating some of the downtime. Because some of the, you're, you're eliminating some of the decompression that, some, some agents kind of like, if, if it's you're, imagine you're constantly on one nine one one call and

Irina:

Yeah.

John Walter:

next 9 1, 1 call. Then you're on the next 9 1 1 call, and then it's and you like, then compare that to. And, and like AI is just transcribing and just AI is just feeding you 9 1 1 calls as fast as possible, right? compo as compared to Okay. Like between each 9 1 1 call, you have a few minutes to take notes. That's, that, that might be, especially if it's something that's almost like traumatic, some cases it might almost be like therapeutic to'cause one of the ways, like in counseling. In counseling, you process information through speaking By, and that's why talking with friends is so healthy, about talking with, just talking with somebody about your problems is extremely healthy and, and and but I, most, most call center tasks are not that intense. But I, but I've heard that said before though, that when you, when you start automating of the after call work, you, you you can increase intensity.

Irina:

That that's true, yeah. And that's, I think, where the conversation starts. If we automate absolutely everything, then we wouldn't even need humans. Then just let the, the bot handle everything and then we can pay premium if you want, specifically to be handled by a human. Which I guess it's a fair point for me. I just wanna know that I have the option because I'm that type of a customer. I want someone to pay me attention. But, but it's absolutely fascinating because I think we're entering the stage where laws have to be changed for the adoption of AI. But how fast does that happen?

John Walter:

Yeah, it's gonna be slow. Right now, all of the, the laws, there's a ton of laws being implicated, existing laws I. Need watch out for today are the laws that were enacted in 1989, and everyone's forgotten about, and there's been no litigation on them in over a decade. And all of a sudden some creative plaintiff's lawyer is dusting off the books and, and say, ah, look at this one, look at this one. That's where, that's where you see a lot of the activity today, that the dust will settle around those over time. They're going to, all of these different theories will be tested and, and companies like the one out of Illinois, we're not getting a lot of legal clarity because all those cases are settling because the stakes are so high. But so they're not coming to a conclusion. There's no finality. We don't have any clear, we don't, there's no clarity from the judges on what's allowed and what's not allowed. So company, but companies are learning how to avoid circumstance of getting caught in that trap. And so as the dust settles, we're gonna have a proliferation of state laws, state, at least in the United States. I'm not, and I'm not sure how it works in Europe, honestly, when it comes to the European Union. If you have individual countries that have this, is it like the, is that kinda like the US where you have like France might have a specific law on, on regulating employment and then you have the European Union also having,

Irina:

Yep. Yep. So it, it's gonna take a while but before we wrap up. Any last words of wisdom, passion about AI? I, I think we might have to continue that conversation. And one thing that before we end up, I want you to, to mention is how do people find the AI association, how they get in touch with you, how do they find you?

John Walter:

Yeah. Yeah. The, the, the last words of wisdom I'll just share is there, there's a t tremendous when it comes to just sharing the pulse on the, the industry like if, if what if, what if, what context inter leaders are expressing are the most interest or most pressing concerns to them today. And just in case other people are sharing the same things, to know where you can find that community is Over the past few months it's been. The deployment of voice AI, there's a big gap in the industry among the, the number of contact center leaders who have deployed voice AI in, which is so fascinating. In the past, a lot of this technology stuff has been deployed by IT departments or by the self-help departments. self-service, department of self-help departments and the. You know, generative AI, conversational, generative AI voice is better managed by the contact center leaders that are managing voice conversations all day long. This is really with more so in their domain. All of these tools are becoming so user friendly and, and that, that you don't need to be an IT professional. It's actually more suitable to be a contact center leader or, or, or someone who's in, in charge of hand. Overseeing human voice conversations and managing human voice conversations. And, but the, but the gap between the number of contact center leaders who have interest in voice AI and the ones who've, especially generative AI. And the ones who have deployed it is massive, massive, massive, massive. So that's where over the next probably few quarters, we're gonna be focusing a lot of our energy and attention on helping find the, we're identifying the, the contact center leaders who have developed that experience in their own organizations and having them share their insights and lessons learned with others who are now exploring that path. And then secondly, the it is more fresh, but it, it is, it is, it is starting to be the requests from AI and I promise I did not instigate that conversation. Inside the association, I try to keep a very clear, I. Bifurcation between my role in the association and at ProxyLink, but that has become the, the, the second largest topic of interest is, is how, what do we do? My, my, my agents are getting calls from AI and, and what do we do about this? Even, even just today, I, I, I saw a post, you know about this on LinkedIn, on, on the AI vendor side providing tips. On how to avoid your AI from being flagged as AI, it's getting to a point where a lot of call centers are having a difficult time even distinguishing human from AI communication. And so, and so that's, that's another big topic if anyone's interested in just networking with, and, and sharing insights and experiences and building friendships with other leaders and who are similarly situated. the Contact center AI Association's website is CCAI.org C for contact center AI association, so ccaia.org And then, and then for, for ProxyLink, which is the, the service that I, I've, I, I started last year, which solely focuses on authenticating inbound AI. Calls and, and chats and all that good stuff it's ProxyLink pro ProxyLink.co. And I'm, I'm active on LinkedIn and so just John, there's a lot of John Walters in the world, but but yeah, John Walter on LinkedIn,

Irina:

Perfect. Thank you so much for that, John. And I'm actually very excited and very intrigued for joining that association because what you were mentioning in the conversation that we had before we kicked off the recording, I told you. I hate the topic of AI when people cannot give me examples and you've been nothing but giving example after example, which is exactly what can make or break how, how we use technology or whether we use technology. So I'm definitely gonna continue that conversation. How, how can I.

John Walter:

for you to get involved. Let's do it. Let's get, let's do it. Awesome. Thank you for taking the time to talk. It's always a joy to catch up. I love seeing what you're working on and, and let's, anytime. If you ever want any guests, one of the things that the association is starting to do is there's a lot of different industry organizations. That are constantly seeking contact center leaders who have experience and are willing to speak on these topics, and we're starting to help them to find speakers. And so if you ever want someone on your podcast who has experience deploying this stuff in their own operations, I'd be happy to bring someone from the association to help share their insights and, and, and we would love to get you involved to help share your lessons. Learned also.

Irina:

Perfect. Thank you so much. Let's, let's arrange it and chat soon. You're always welcome on the show.

People on this episode

Head Shot

Irina Hollatz (Mateeva)

Host

Podcasts we love

Check out these other fine podcasts recommended by us, not an algorithm.