Name: AI and Labor Markets: What the Data - and Leaders - Are Really Seeing
Uploaded: 2026-05-20
Duration: 18 min 31 s

AI is reshaping global labor markets faster than most organizations can track—but does the reality on the ground match the hype? The discussion offers a bird’s-eye view of today’s landscape—from shifting demand for entry-level jobs to how employees are adapting through automation, reskilling, and, in some cases, layoffs. We will pair real-time labor market data with firsthand executive experience to address the defining question of our time: What is AI doing to the workforce right now—and what comes next?

Svenja Gudell, Chief Economist, Indeed
Becky Schmitt, Chief People Officer, PepsiCo
Moderator: Ruth Umoh, C-Suite and Leadership Editor, Fortune

AI and Labor Markets: What the Data - and Leaders - Are Really Seeing

Conferences

Thank you all for being here. Uh, AI is moving faster than most organizations can fully track, and the conversation around work has become incredibly polarized, to say the least. What I hope we can do today is ground this discussion and what is actually happening inside companies and across the labor market right now. Let me throw first level set, throw those questions to both of you. It seems to me, and again unfortunately we cover this ad nauseam, every AI conversation seems to swing between two extremes mass job loss or total hype. What's the single biggest misconception about what AI is actually doing right now? You know, from, from our perspective, we, we sit on a lot of data at Indeed. We have an incredible perch. We see the entire marketplace, so we're able to see demands that employers have for jobs and of course, uh, the job seeker side as well and what skills and experiences job seekers bring to the table. So and how, how the two interact. So it's a lot of data and this question comes up a lot of. you know, is AI destroying all sorts of jobs, or is it really all hype, and at the end of the day it's obviously the world in the middle, right, because we, our own research actually shows that AI will touch every single job out there, you know, every job will be touched by AI, some more so than others, but at least currently with today's technology, not a single job can be. Completely done by AI you still need the human in the loop, and I think having that recognition is super important and then, um, we are our research actually points to the fact that we will certainly see some jobs be very impacted by AI and we're seeing a little bit of that in the data already if you're a software developer, uh, you know, you, you are having a harder time finding a job right now um however. We're also seeing that AI is creating a whole bunch of new jobs and interestingly enough if you are an AI software developer, uh, things are looking quite good for you. There's a premium, a wage premium on your skills right now, and our entire sector of software development is actually up for job postings 14% year over year, uh, which is, you know, surprising to a lot of people because they're like, oh, that must be a hard hit sector. And the sectors that are most exposed to AI right now are seeing the most growth in terms of demand for those jobs, so there's a lot going on and there's no question that AI will transform a lot of things. I will leave you with one thing before I hand it over to Becky, but it's really important to. Also recognize that there's a lot of stuff happening right now. It's not just AI, and demographic change is a huge one that's knocking on our on our door right now, and we'll see a lot of impact. And when you say demographic change, you mean what exactly? We have an aging population in the US and really most industrialized countries are seeing an aging population. Uh, so we have baby boomers, you know, continuing to retire, and that's a huge generation, right? So we see a lot of jobs being impacted by that, and the younger generations aren't as large, and so we're entering really a new phase of really a great mismatch because we're going to have workers in places we actually don't want workers and we're going to have to reallocate a whole bunch of jobs to other people and Um, that's gonna cause a little bit of pain. We're gonna talk about that reallocation with more granular in a moment. Becky, I would love to hear from you. Yeah, well, first of all, thank you for having me, and I would, uh, echo everything that was just said, but for a global multinational 100 year old company, these challenges are very real, and a lot of, uh, when you see, oh, jobs are going away, I think right now there's just a lot of basic efficiency that's happening, and, uh, we can still. You know, derive a lot of, uh, process standardization, cleaning your data that creates more capacity in the system and I think for us at PepsiCo we have some jobs that people don't want to do anymore. They're very manual. It's hot outside. It's longer hours, so how are we using technology which may or may not be AI at this point. Sometimes it is, sometimes it isn't. Give us a for instance if you could, uh, so a couple of years ago. In our Mexico business we established a control tower so we deliver to over 800,000 retail outlets which could be anything from a restaurant to a bodega, what have you and we've got all these drivers on the road uh if there's unrest, weather patterns, construction, how are we monitoring what's happening in that vehicle at all times so we can either deploy help or reroute the drivers now that practice which is actually creates a really safe. Environment which is attractive to people who have those jobs. Now we are underpinning that with AI so your drivers are not going to know oh now AI is helping us be safer, but actually that can synthesize at a much faster rate what's happening through news channels, traffic reports, etc. and do dynamic routing for our drivers as well as make sure they have the safest, most, most efficient route. So in many cases what started. As process standardization using a common set of data standard automation now you underpin that with AI and I think you create velocity in the system and that helps us deliver more predictable, make the proper amount of products, as well as streamline some of the harder jobs that because of the demographic changes people are just less interested and we're even finding in countries where you still have large populations and a lot of young people. You know they wanna be influencers they don't actually wanna be a sales person and so how do you say, well, you're gonna influence our customers, right? So there is this mismatch also in interest and so we've got to be more efficient. Let me follow up with you there, Becky, because you sit inside a company that is quite large, uh, quite operationally complex, over 300. 1000 employees if I'm not mistaken, spanning some 200 companies. Does this moment, this AI era internally feel like a true workforce revolution, or does it feel like yet another layer in a long running productivity agenda because I hear time and time again enhanced productivity. Does it feel markedly different to you? Um, I think it feels different in a sense that there is more pressure to explain things to the workforce because people are getting a lot of information and I mean even me every day I have another group that wants to talk to me or help me plan the effects of AI. I mean we still have legacy technologies. We have 50 year old factories. Things still need to run, but again, like how do I make jobs safer, better. more productive and more attractive and that's the discussion we keep having is like can we create transparency for our people? How do we use a human centric design? what is the accountability and how do we audit these new technologies so that they're doing what we said they would do and they provide value uh because we do believe that humans are going to create new opportunities and we believe in reimagination. And I think that's going to come from our people, not just from the tech and people using tech in the right way. So that's what we keep talking about. And even when we roll out simple things, you need adoption for it to work. Uh, and there are some places where we have not actually had a lot of technology innovation in certain countries, and they're like, where's mine? So I think it's a mix in the middle. Uh, some geographies are getting more focused than others just based on like the density of our business, but people are looking for a change because in your personal life we're all using this all day long through our devices. So why would you come to work and fill out paper, right? It feels like we should have a change. Let's ground this in additional labor market data, Sonia. I wanna turn to you. Indeed recently published new research, and your data suggests that AI is transforming tasks more than it is eliminating whole occupations. But at what point does task erosion become job erosion? I think there's a lot in that, and interestingly enough, our research also shows that we're In the very early days of all of this, right, every, right, and every stage will look a little bit different, right? Overall, um, our AI tracker actually shows that only 5% of job postings right now have an AI mention in them. So there are a lot of jobs that don't mention AI at all that probably don't use AI at all yet. So it's still very early days. And then, um, on the skill side of things when you're actually looking at how jobs are being transformed, we're seeing a lot. Of the normal skills that for example you know take a program manager, they have to have all the skills that they used to have plus AI skills so we're layering AI right now and to me that is also a sign that it's still very early on because we're not actually you know changing whole processes quite yet. We haven't gotten really good at infusing AI into the process. You probably have to change the process a whole lot to make it work well with AI and I'm sure you're seeing this on your side as well. Where you, you can't do it the way you did it yesterday, you have to do it differently, uh, but we haven't quite gotten there yet, so it's currently still a layer on top of that. And then you're absolutely right, there's going to be a rearranging, right? It's perhaps if it makes you more productive, a company will not need to hire 5 people for that particular task. They'll need to hire 3 people, right? So there is, there's a. Little bit of job loss potentially and something like that, but, uh, it also makes you more productive. It reduces costs potentially, although yeah, it's kind of expensive right now, but uh you know, in the, the, the future will hopefully hold some cost reductions and then um you can do bigger, better things, and then maybe you will hire 7 people. So all these things kind of play out and there are lots. The forks in the road and at every fork you have to decide and it'll be interesting to see what fork every business will take. I mean, AI without a doubt it's dominating the public conversation. It's certainly dominating conversations among HR leaders, but I noticed indeed data it shows that AI hiring is still highly concentrated. Um, AI jobs remain a relatively small share of total postings, a relatively small. Number of firms accounts for most of that demand. Are we overestimating how broadly AI transformation is actually spread across the labor market? I think we're overestimating the speed at which we're going to see this transformation. I think we're underestimating the long term impact it will have, right? I think it's going to significantly change the way we do things, um. But not, not tomorrow or the day after, but months and years into the future. So I think that's, that's a really big driver and, uh, for all of you in the room, right, that change happens with you too, right? How you integrate all of this into your processes and that's why I think it's so important to kind of have the data and to be able to act now to kind of kick off the things that you need to do in order to be set up for tomorrow. Um, a difficult task for sure, but something that's really important. Certainly. Let me turn over to you, Becky. Uh, you earlier in this conversation alluded to something that set off something in my brain. You said AI infrastructure compete investment. They're concentrated in a relatively small number of countries, certainly also companies, but absolutely countries. How does that reshape the global map of talents, let's say at PepsiCo over the next 5 years, and what is Company like PepsiCo need to do to avoid creating uneven opportunities across regions. OK, this is a problem I don't have an answer to. So anyone listening today, I welcome the dialogue, but I think about this a lot, and I thought about it before the impact of AI. Sometimes when you have a company that exists in over 200 countries and territories, you end up having haves and have-nots, right? Some factories get. More efficiency or they're producing a product line that's more profitable so it drives greater innovation versus some markets well the cost of people and technology is disadvantaged, so you let it be but really our people around the world are asking what about me? You know, I'm seeing in the news or I'm hearing about this happening in one part of our business. When are we going to see it? And I do think about like there is one stock price. We have one CEO and we do try to have a common employee value proposition and part of our success has been moving people around the world so no one's going to want to go from a really advanced place to a place that's not getting the investment unless they're there to go make it happen so I don't have an answer for this, but I do think how do you manage a big organization when one part's moving faster than the other. And that's something we're working on internally right now. And what pieces do you invest in one place that are very scalable and the what, where have you invested that's quite scalable? Uh, well, we, there's a lot that we do in the Americas. And when I talk about the Americas, I talk up and down this part of the world just because that's, uh, 80% of our revenue, right? So it gets that investment. But the way that you make it scalable is by having standard processes, accepting common data sets, having core parts of your job that have. You know, one job in one country and another be translatable, so then then the technology can move. But if you want to let everybody have their own kingdom, you're going to invest heavy in one part of the world and then it can't go someplace else. And we see this even in product innovation. Like if you like baked lays, how do we make sure that technology for oven baked can scale? I know that's not something we're thinking about when we talk about AI, but it's the same concept. How do we move that across the world at scale? Follow up on the Pepsi. Go thread. One thing that your employer is known for is being a Fortune 500 CEO Minton factory. You have quite a stringent and deliberate process for sourcing and tap in high potential employees and developing and training them to be leaders internally and certainly externally as we've seen time and time again. How does AI change what high potential means inside an organization? Are there certain skill sets that are increasingly requisite? Or conversely have atrophied in the AI era so it's a great privilege to be able to be at PepsiCo with this sort of history, but it's also, uh, a challenge for me to evolve what it means to have a dynamic profile that continues to generate great leaders whether they can they stay with us or they move to other organizations. I think there's a couple things that are part of our secret sauce, and I'm sure a lot of companies would say this, but our people have hustle. And we have to be able to identify the hustle, uh, problem solving, which I think is a key, uh, component to being agile for the future and whatever jobs come up. How do you solve problems? How do you stay like have that internal fortitude to work through it? How are you curious? How are you always asking questions just because you've had a process that's been working for 40 years. Years why would we keep doing it that way? It can be different. What's stopping us from the future? So we are actually have a big project underway right now. It's called like GM, general manager of the Future, and we're looking at all the data we've collected for 20 years. What makes people successful? Did their data, their assessment, was it their lead indicator or actually was it the jobs we put them in? Was it the products they innovated? Was the problem they solved? And how do I reformulate that so that I can have a deep bench and so we've got a whole team working on it and I think that it's really about, you know, what's the person's education, the one they come with, the one we provide, what exposure can we give them, how do we look at on the job experience and then how are we modifying our evaluation process so we're finding these jewels. Wherever they sit in the organization, it doesn't just mean the next line down, but like deep in the organization and we rotate them. So it's like it's the kind of work all HR professionals really wanna do. Absolutely. Let me turn over to Yesenia because one of the most consequential findings in your data is what's happening at the entry level, especially in sectors like tech if AI is absorbing a lot of the routine work that junior employees have historically learned from. Are we quietly breaking the bottom rung of the career ladder? Where do experienced employees come from 5 years, 10 years down the line? You bring up a really interesting, um, point because we do sometimes hear employers saying, I don't hire entry level anymore, right? And I would like to hire only people that have 3 to 5 years of experience and we see this in the, in the data where we actually have seen a pickup in job postings requiring. More experience within the tech sector, uh, this doesn't happen so much in other sectors, but, um, it does break your pipeline, right? And, uh, in 3 to 5 years from now you're not gonna have the people with 3 to 5 years of experience because you never hired them to begin with, right? And so, um, that is a distinct pipeline problem that we're all gonna have to bridge and I think that means what it means to have an entry level job is going to change. The job itself has, has to adapt and, and change. And um obviously you know I always think AI is not going to take your job, but the person that knows how to use AI productively will, right? So how can you infuse AI how can you infuse technology and meet the needs of this, this new job, uh, with, with these different task skills and whatnot and then of course schooling has to change and the entire roadmap into, you know. From, from growing up into the into a job we have to adapt, right? And so that works will have to be done. I mean it's an entirely different conversation, but I certainly do think that one of the key fears is that we're training employees for yesterday's economy, to your point, um, but this has been such an insightful and honest conversation. So thank you both for what has really been a robust discussion. Thank you so much.