HR's Role in AI-Driven Change

[00:00:00] So most companies are still figuring out their AI policy and what execution on AI even looks like. There's a smaller group that's actually done it already and they're using it. So today we're going to do lessons learned from someone who's been part of that crowd that's already done it. And I'm Andrea Adams, this is The HR Hub and today we're talking with Laura Maffucci. She's Chief People Officer at Globalization Partners.

[00:00:26] And GP is a global employment platform that helps companies hire internationally and quickly. Laura's been in the middle of this AI implementation and today she's going to share what they've done, what worked, what didn't, what HR really needs to know to help lead through this. Hi Laura, how are you? Hi there, I'm good, thanks. How are you? I'm great. I'm great and looking forward to this. You know, so I got on the bed when I get in early about like, you know, what is AI?

[00:00:56] What is AI? What does it mean to HR? But this is the first lesson learned I've done. So we have, we have gotten through that. So I'm looking forward to this. So can you like delay the groundwork for us? Can you tell us about the scope of the AI implementation at GP? Sure. So we're actually just starting a new phase of it too. Oh.

[00:01:20] So we initially, you know, the mandate came down, everyone should use AI in their jobs. And at that point we already had an AI council, policy council in place that our chief information officer had rallied some people together for. So it was myself, a deputy general counsel, some folks from the AI team, our head of AI,

[00:01:43] AI as well as, you know, systems, IT security, that sort of thing to really have a policy in place for usage. We already had that and we were fine. We had had it for probably about six to nine months before the mandate came down. And we were really focused on just keeping the policy up to date because they knew tools were coming out. We were trying to keep a list of tools that were okay. And like, it was just moving so fast. We found that we were meeting, you know, every other week and updating it monthly.

[00:02:12] And then that happened, we got the everyone should use AI and quickly pivoted into, okay, how do we enable people to experiment and make this a positive and not make them scared about it impacting their jobs? Because I think the first thing everyone thinks is, oh, they want to replace people with AI. Am I being asked to use AI to work myself out of a job? So you want to tamp down that fear.

[00:02:37] So quickly set up a Slack channel for people to share the ways they're using AI, which has been really cool. They're sharing how they use it for work, but also fun things they do with it in their personal lives. We did a bunch of, and we're still doing on a regular cadence, lunch and learn, so to speak, or learning sessions where people can join. And we have folks from our AI team there to answer people's questions about AI.

[00:03:02] We brought on an enterprise level tool for people to use because we didn't want people just out there putting our proprietary information into chat GPT, which we were finding people were doing. Yeah. We brought that on and it wasn't chat GPT. And we were getting, you know, a lot of complaints. It's not as good. It's this, it's that. And it really turned out to be prompting errors more than anything else. People just don't know the right ways to prompt.

[00:03:25] So luckily being a technology company and building our own AI product, we have very smart people who know how to help with that. So, you know, we did some prompting classes. We also have an AI awesomeness awards to kind of encourage that experimentation. And we've had huge outpouring for that.

[00:03:48] And the fun fact that our CIO is most proud of is 70, I think it's 77% of the people who have won those awards are not on the technical teams. So the whole organization has really gotten out there and have been experimenting and building some automation.

[00:04:07] But now we're just about, we're just kicked off a new phase, which is a more structured approach to really reimagining work and really in changing the way we work and embedding it into processes where it makes sense to have the automation. And so, and that's going to take us in a whole different direction, which is really exciting.

[00:04:31] Okay. So to recap, the first, the phase that you've just mostly talked about was really encouraging employees to use it on a daily basis and to play with it and to experiment with it. Because they weren't even doing, they weren't doing that without that extra level of encouragement. I mean, some work, you know, but I think it was more just like, oh, let's use chat GPT or can I try notebook LM? Like there was, there was little things like that.

[00:05:00] But it was really to encourage them. And we, you know, we put cadence, we put a processes in place so they could easily get the tools approved because of course you want to make sure it's approved tools, the security checks out and all of that. So, you know, it was really, it was unstructured. It wasn't really this, a whole department looking at its workflows to say, how can we find efficiencies? It was really grassroots, like get people just using AI.

[00:05:24] Because at the time, that's just what you did was everybody use AI, you know, it wasn't, I don't think anyone had really started looking at a structured way of doing it yet. Right. So now the structured piece that, I mean, I'm interested in, in the unstructured part because I think a lot of organizations are still like, okay, how do I make this go? But the, but I'm also interested in the structured part of it. Okay.

[00:05:53] I'm, I'm, that's where my brain is right now. So I'm going to ask what, when you say structured, what does that even mean? So, and I, I, the full credit goes to Layla, our, our CIO. She is, this is her superpower, transformations like this. And so she, she is leading along with the CFO is leading an initiative to, like I mentioned, reimagine work. But we're going broad in some areas and we're going deep in others.

[00:06:21] So in some areas like in HR and in legal, it's more about, okay, let's identify some workflows that could be automated, that don't necessarily, that are repetitive tasks, easily automatable, don't require a person do things that it makes sense for a machine to do type of thing. Right. And then there's bigger pieces to it that is really going to go deep in areas like marketing and sales and the whole go to market engine and the lead flow and, and how we can,

[00:06:51] find automation there. Because just imagine if you have a Salesforce that can generate a certain amount of revenue, if every single one of them had like a little AI sales buddy that was running, you know, doing similar things and out there working for them, could you double your, your revenue because you have that extra AI help. So it's really about finding more efficiency and ways to grow without having, adding humans to grow. So it's not about replacing humans.

[00:07:19] It's about making them more productive, keeping that human in the loop and giving things to help them be more efficient and more effective in their jobs. So we're going to, we're working in a way that every, every function has a person they've named as their lead for it. And, you know, some of the bigger functions, they have several. Um, and where everybody gets an AI advisor.

[00:07:43] So we're bringing in consultants or contractors who are experts in AI and know how to get agentic and know how to build these agents, um, to work with them and really be that, you know, when you, one thing I've learned from working at GP and working with the product team, when you work with a product team around, or you work with somebody who's helping you with technology, which I never knew how to put this into words, even though I, I innately felt this having been in HR, my whole HR is my whole career. Tell them the problem you have and let them solve it.

[00:08:12] Don't tell them what you think. Oh, I need you to build me this, this, and this. You're going to get a much better product. And so we don't know anything. You know, I, I have some, it's some AI geeks on my team who are pretty into it, but I don't think any of them feel they could go out and build an agent, but they know what problems they think could be, and processes could be automated. So just saying, here's how we work and here's what we do.

[00:08:34] Here's where the pain points are and let an expert tell us and help us build what we need to solve those problems versus I think a lot of companies are just bolting on these existing solutions. I mean, every company out there, every time, every software company was like, Oh, we have AI, we have AI. And people are just going out there and buying it. They're slapping it on to existing processes. It's not really driving anything. The way to really get good results and an ROI from it is to reimagine how work is getting done.

[00:09:04] Look at it and say, if I was starting this function or this process today with the tools I have today and the tools that might be different in 45 minutes or in a week, how would I, how would I build this today and do it from there? And that's how you're going to find AI solutions that are actually meaningful and not just AI for the sake of AI, which I think a lot of people are doing. Yeah. Yeah. Okay. Can you give us an example? Oh, boy.

[00:09:32] You know, I just want to, for the audience, articulate the difference between, in an example form of just sort of bolting it on versus rethinking the whole thing in a way that is more efficient and sort of integrates AI and how it gets done. Well, I think, I'm not sure I have a specific example.

[00:09:56] What I would say is if you are taking, you know, say you use Salesforce or you use Workday or you use any of these platforms, they have an AI component to it. And typically you're just going to like adopt that and use that within the ways you've always been working. When you start to go over into the agentic space and having agents do workflows, that's a very different thing than just taking advantage of that little solution that comes with something.

[00:10:23] You know, with our own product, GIA, which is a great solution for people for compliance and legal and employment compliance and things like that. You know, we're helping the team and we're going to do our own experiment of pulling it in to see how it can help us actually with some of our workflows to help make that more agentic. But to do that, we're taking a step back. Yep.

[00:10:44] So around GIA, you like, and I'm sure compliance is just huge in what you're doing as an organization because of it's so cross jurisdictional. Um, but it's kind of a bolt on where you go to it and ask questions versus what you're contemplating, uh, structurally is where it's part of the flow. It's doing things without ever being asked. And GIA is doing things without ever being asked.

[00:11:14] It's just, we're still in the early phases of it. So I wouldn't consider that a bolt on because you really would have to take a step back. I mean, you, you, you can upload all of your policies and it can tell you when it needs updates. It can generate those updates automatically without you having to prompt it at all. So it's, it's got an agentic flow in it. It's agentic. Um, it's very purpose built. It is based off of our proprietary in-house knowledge of over, you know, a decade of employing people around the world.

[00:11:43] It does scrape the internet, but not the way the other sites do. It scrapes government websites for regulatory updates. It's not getting answers for your compliance questions or to build your employment contract for you or to do your policy review and update. It's not getting that information from Reddit or Quora or the other places that a lot of the large language models, that's where most of their answers come from. It's just general websites and you have no way of validating it.

[00:12:11] So all of our data is GP, you know, we have a badge right there. It'll tell you it's GP verified. You can click on it. It tells you the source of the information. Um, if it is coming from a government website, it'll link you right to that. If it's internal, it brings you to the knowledge document. It's very, it's very purpose driven. So, um, I think I wouldn't consider that as much, as much of a bolt on as just some of these, you know, like zoom has the AI assistant.

[00:12:35] So you start using AI assistant, but you're not really rethinking how, how you're doing, you're doing things. You know, with Gemini, you have the gems that you can go in and build. And what we've found was so many people experimenting, they're experimenting and they're building for their building solutions for their, the process as it stands today.

[00:12:58] And we're finding it's being adopted so widely that we almost are running into the opposite problem around encouraging use where now we're trying to take a step back and do it more structurally. We're worried about, you know, the shadow AI taking over and all the different people doing things we don't even know about. And it might be working in ways that, well, we were actually trying to shift and maybe not work in that. We were going to change that way of working and they don't know that.

[00:13:21] And, you know, so we're trying to make it more mindful and really take a step back and reimagine it. We're working differently now. So it's not about taking your existing processes and recreating them with AI. AI, it's about looking at the work that needs to be done and creating AI based processes, really. Right. Okay.

[00:13:48] I have to, I like, I have to get to the, you know, you being HR and HR's role and how you see HR assisting and everything. But before I get there, I want to talk about the AI Council. I think you're a part of it. Can you just tell me about that and what it does? Yeah. So like I mentioned, we drafted the original policy and it's something that we review on a regular basis and make sure that we keep it up to date. Everything changes so quickly. You want to make sure you have things covered in there.

[00:14:18] So there's that. And we've spent a lot of time since we've done all these initiatives to encourage the usage. We approve requests for AI tools. Okay. So it's a way of having that governance to make sure that, you know, if this team's already got this one AI tool, will it actually help what this other team is trying to do as well as opposed to going out and purchasing two of almost the exact same thing.

[00:14:43] So getting a little governance over what's being brought in and the security aspect to that as well. You know, we brainstorm around, you know, how to get the education out to the employees like those learning sessions I talked about. All of that runs through the AI. All of that runs through the AI Council. Like you've been pretty open with AI and what it sounds like in any ways.

[00:15:08] And the tools people are allowed to bring in and use, I think that scares a lot of companies. What do you think is the right approach to AI policy? Well, I think you have to create a positive energy around it. I think that's one of the things that we were able to successfully do, which, you know, you never know if people started adopting it at the pace that they did in our organization because they were afraid. And then it led to them using it like the fear may have driven the good behavior.

[00:15:35] But I think that it it's dissipated and people are genuinely enthusiastic about learning it. So you want to make sure you're creating that environment. And, you know, you want your policy to encourage the usage and balance that with providing the governance to it from and mostly from a security standpoint. So I think you want to have solid governance, but also combined with really encouraging that experimentation.

[00:16:01] Don't, you know, don't have it be a thing where somebody is afraid to, you know, produce a work product and hide the fact that they use AI. You should be able to say, oh, I used AI to help me do this. Right. I was going to ask about that. There's a stat out there. It's somewhere in the range of, I don't remember, 67%. And of course, this number is going to be changing all the time. But the majority of people hide their AI use because they want to seem more valuable.

[00:16:29] They are afraid of AI taking their job. So they don't want to confess to how much or they just want maybe the pace of work and change and everything. They just want a break. How are you getting employees? I think it's kind of, it's perceived incorrectly. You know, one of the things I've been looking at a lot is that gap between how, I think I've seen it called executive delusion.

[00:16:53] That gap between what the boardroom and the C-suite think AI is actually doing versus the value that people are actually feeling on the ground, the employees that are using it. And a lot of employees are using AI and then they're, you know, producing that work product to someone who they report to. And it's real with errors because they're not applying their own critical thinking. They're not, they're not, they're not remembering like a human needs to be in the loop. AI makes mistakes. AI makes a lot of mistakes.

[00:17:23] You have to have that human oversight to make sure. I mean, I think there was a, I saw something, there was a company that couldn't get anyone to use AI. So then they spent all this time building an internal system with all their internal data and still no one was using it. So then they mandated people use it. Well, then everyone started using it and they found out like a month later that all of the sales numbers that it was generating that people were using were wrong. So you need that, that human oversight.

[00:17:50] So it's, it's causing some people more work because then you're having to go back to the person or you're having to redo something that somebody has given you that was AI slot or you're spending a lot of time yourself. It's not necessarily always a huge time saver. Yeah. It may be a little more efficient, but you're spending so much time going back and forth with the AI. It's just what you're doing in relation to it is different. It's almost like having an intern or like a very junior employee doing work for you.

[00:18:19] And sometimes that can make your job a little bit harder, even though they're there to help you and take workload off of you. You have to spend so much time coaching them and training them and reworking what they do. It's sometimes you're like, it would just be easier if I didn't even have them or if I was doing it myself. I think that there's that, that feeling out there for people.

[00:18:36] So yeah, I've even had that experience now where standards are going up because like they did when, you know, the advent of computers is that are not computers, but say, well, maybe it was, I don't know. Anyways, like because you can get a better quality product, expectations rise, you know, your, your written product can be, so you're getting more revision.

[00:19:02] So it is both making what we used to do easier, but now we're having to do that level quality of, you know, written whatever, all the time. And so it is taking just as long as it always did to get to the final result. Yeah, it's just, it's just a little, it's just a little different. Yeah. Yeah. Now that we have to talk about HR, of course, this is an HR YouTube channel. What's HR's role?

[00:19:29] Where, where do you see HR adding the most value? Um, the reality check, the, the change management and the, um, the, the keeping it real. Um, you have to be willing to speak up about what employees are actually experiencing with AI.

[00:19:50] You have to be willing to speak out up about the fear that exists there and help leadership message it in a way that, that is authentic and isn't ignoring what their reality is. And I think that's a mistake that a lot of companies are making is they're like AI, AI, AI, this is wonderful. But they don't, they don't even know, they don't even know what they're expecting employees to do with it.

[00:20:17] They're just telling them to go and use it or they're saying, oh, we got all of this efficiency. And the employees are like, no, I didn't get that efficiency or it's not, or, or I want to use AI. You're telling me to use AI. You're almost faulting me for not, but I don't know how I need help. I need someone who knows how to build this to help me build this. And I think that HR plays a critical role in being courageous and making sure that you're saying, wait a minute, if you say it this way, they're going to feel this way.

[00:20:45] Make sure you're, you're, you know, you're coming at it from this angle. And, you know, I think, um, and I'm very open about this. I hate AI. I think it's the end of civilization as we know it. Um, but I also am, you know, let go or be dragged and I know it's there. So you have to make friends with it. You have to learn how to use it. And there are certain things that it's good for and that it can do.

[00:21:08] Um, so you, you have to stay very, um, mindful that a lot of people, a lot of people feel that way. And I am very proud of the, and it's a lot for me to say this because of how I actually feel about AI. I'm very proud about how we're handling this.

[00:21:28] Um, you know, from even, even when we were just starting and encouraging it with the IA council, but especially now with this bigger project that we're undertaking, we're doing it in the right way. We're doing it for the right reasons. We're doing it with the right mindset. We're being mindful. We're getting people to support to do it versus just expecting them to figure out how to do this on their own. Um, and we're involving all the right people. And I just think I am really proud of how we're doing it.

[00:21:56] So, um, I'm lucky that I'm working in a company where I can play that role. I don't know that that's so easy for everyone in HR at every organization. I know a lot of organizations that aren't tech organizations. I get questions a lot, like out at conferences and stuff, you know, about, well, I don't have anyone in house who knows anything about it. Like how do I even get started? And what I always tell people is there's people in your organization who in their personal life love it. And they're in there, they're playing with it.

[00:22:24] Bring them in as champions of it and start that way. Um, so, so yeah, I think it's, it's, it's keeping it real and being on change management really. Yeah. Yeah. I just have to acknowledge, um, your dichotomy. I don't know if that's the right word, but the, your internal gap between hating it and embracing it or, I don't know.

[00:22:51] And, um, I guess kudos both to you too for acknowledging that, but also persevering through that. I don't know. It's exhausting a little bit. I'm not going to lie. It's, it's not, you know, but what, what is good is, you know, I, I mean everything I'm saying. I can, I can sit here with you and I can talk about the good and the positive things about it, but I'm also able to say how I really feel about it because it doesn't cancel out. I do see the efficiencies.

[00:23:21] I use it. There's, there's things I use it for all the time, all the time at work, use it all the time. You know, most I'm, I'm still, you know, I'm Gen X, but I'm kind of a boomer when it comes to AI. I kind of treat it like Google. I know I don't prompt it right, but you know, I, I use it. But there's certain things I would never do. There's certain things that I think just take it too far. There's, there's, there's things about it that I'm like, I, I, I roll, you know?

[00:23:45] And I do, I have a real issue with the approach a lot of companies are taking where they just, they don't, they don't know what they're really asking people to do. They don't even know what they want AI to do. They're just telling people they have to do it. And I think that is the hardest thing in HR is you're being asked to lead the change on something where no one's getting any resources to do what needs to be done. And that's where it gets really hard. And that's where I say, I'm lucky.

[00:24:11] I don't know that I could do this if I, this would be really hard if I couldn't sit here and say, I'm proud of how we're doing it. Yeah. So I have the luxury of working somewhere that's doing it right. What does HR do with, you know? So I think it's about, and maybe I'm wrong, but I wholeheartedly believe that it's, it's about being authentic.

[00:24:34] I think that people need to hear that people hate, there are people who hate AI, hate AI, but you're still, you're still going to adopt it because, hey, it can, it can do these good things. People are tired of corporate speak jargon, just blah, blah, blah. It doesn't mean anything like they're saying they're going to do this. They don't know what they're talking about. What does that even mean? It's just a buzzword. They're tired of all that.

[00:24:58] And because they're questioning if everything is real, having that authenticity, I think, is the most important thing. Don't just be the corporate talking head. You can acknowledge being a real person and having a real opinion about something that might not be, you know, 100% the, you know, the corporate vision statement on the wall. You can do that and still be an excellent employee champion for the company while also being a genuine person with the employees.

[00:25:28] Where can someone find you and find GP? Well, we have a huge link. I'm on LinkedIn. Yeah. And we have a huge, GP has a huge presence on LinkedIn. And our website, GP.com, is also a great place for lots of information. We're also, you know, anyone attending any of the conferences, we almost always have a very solid presence at like SHRM. I actually will be speaking at SHRM and with our chief product officer.

[00:25:57] And so any of the HR conferences, you can typically find a GP booth there. Well, thanks, Laura. That was great. I really appreciate your authenticity as you talked about in and being so open, you know, both supporting it, but also questioning it. And I like that voice in the dialogue. I think that's great. Um, I've done other episodes on AI.

[00:26:24] It's been a while, but here's one there and here's another one there. Thanks for watching out there and we'll see you next time.