AI, Agents, and the Future of Talent

Chris Hoyt: Oh man, we’re back. Welcome to the Recruiting Community Podcast. I am Chris Hoyt, president of CXR, and I’m your host of this podcast, along with Mr. Gerry Crispin, co-founder of Career Crossroads. Gerry, how are you on this warm day?
Gerry Crispin: I’m feeling really good. I’m waiting for the camp out next week, so I’m kind of prepping for disappearing for a while.
Chris Hoyt: Yeah, you go off the grid with your cousins for the camp out. It’s triple digit temperatures now — how many folks? Not only might it be triple digit temperature, but you’ve got triple digit attendees, right? How many family members this year?
Gerry Crispin: Yeah, 100 of us this year, and we’ll take over an entire campground. It’s the 49th year, so most of my cousins have said that after 50 they may give it up.
Chris Hoyt: I think you gotta ride it out. I think you can make it.
Gerry Crispin: Yeah, gotta do it.
Chris Hoyt: All right, well look, for those who are listening and wondering what we’re talking about — we may go off on a tangent for half the episode, but we’ve got a great guest today. On this show we do our best to bring you industry insights and updates in what we call a fun and informative conversation. It’s brought to you by the CXR Career Crossroads community.
I want to talk a little bit, because this comes up in nearly every TA conversation right now. AI is showing up everywhere in TA. It’s in sourcing, it’s in screening, it’s in scheduling, and increasingly there are agents — and this is a new thing that’s hit the space in TA — agents that don’t just recommend, but actually act on behalf of an organization, a function, or an individual. That’s a new curve that’s begun to really become present in talent acquisition.
That’s exactly where it gets interesting, because the same capabilities that make our teams faster are also starting to raise real questions about oversight, fairness, and accountability. Our guest today, Jessica, is going to spend a lot of time on exactly this tension point — helping TA leaders adopt these tools responsibly without slamming the brakes on innovation. How’d I do, Gerry? Did I sum that up?
I think you did good. Thanks a lot. A few things — before we jump in, we’re streaming on YouTube, Facebook, LinkedIn. I don’t really know why we’re on Facebook, but whatever. You can check us out at cxr.work/podcast. You’ll find past episodes, see what’s coming up next, and find hundreds of interviews with TA leaders and practitioners that Gerry and I have done — people doing really interesting work that touches on how we attract and recruit talent, as well as manage and lead global recruiting teams.
On the site you’ll also find easy ways to like and subscribe, and let us know if you want to join the conversation. If you’ve got somebody you think would be a great guest, let us know — we’d love to have them on. And my last reminder before we jump in: this is an ad-free labor of love. Nobody pays to be on this show, and we certainly don’t pay anybody to join us.
Announcer: Welcome to the Recruiting Community Podcast, the go-to channel for talent acquisition leaders and practitioners. This show is brought to you by CXR, a trusted community of thousands, connecting the best minds in the industry to explore topics like attracting, engaging, and retaining top talent. Hosted by Chris Hoyt and Gerry Crispin, we’re thrilled to have you join the conversation.
Chris Hoyt: Jessica, welcome.
Jessica Aguilar: Thanks, Chris and Gerry. Appreciate you having me today.
Chris Hoyt: We’re excited to have you. Jessica, for those who haven’t had the pleasure of meeting you or talking to you on this topic or any other, can you give us that escalator pitch of who Jessica Aguilar is? What does she do, where has she worked, and why is she on today?
Jessica Aguilar: I have a background that’s a little mixed — journalism, computer science, and the arts — and that’s followed me throughout my career. I really love being exposed to different things, different ways of thinking. I started in computer science, then moved into branding and marketing, which was a great entryway into corporate America.
I got hired by a client and worked for an organization where I started in a brand role, and during my tenure there, they had at least three different corporate names. So if you want to go somewhere and get really good branding experience, go through a few large name changes and big acquisitions — you’ll get that.
Through that, I joined HR and had an amazing education in HR over a decade with that organization before coming to the organization I work with today, Zurich Insurance. I spent a few years with Zurich in the US, in the Chicago market, and then had the opportunity to join our group in Zurich, Switzerland. I’m there today, and I currently look after talent acquisition for the group, as well as what we call executive pipeline — talent assessment, succession planning, talent development, executive search, all of those pieces in an integrated fashion.
Chris Hoyt: I love it. Well, I want to jump in and start with the optimistic side of what we’re going to talk about, and I think your background is perfect for this. Jessica, from where you sit, where do you think AI is actually earning its keep in TA today? The places where it’s really delivering real value rather than just demoing well?
Jessica Aguilar: Being able to exhibit capabilities in a demo is a key criteria, but some things work better in display than they do in practice. Where I believe AI is really helping organizations is primarily tied to things that take away from the experience. Think about interview scheduling — that feels like low-hanging fruit. But if you go back historically and think about how much time people spent going back and forth on interview scheduling, and how much frustration was inherent in that process, to me it’s a game changer, because it’s taking people away from something that wasn’t value-added — aside from the interview itself — and moving the focus to pieces that are more value-added, where the human element makes the bigger impact. At a high level, that’s where I see AI playing a role: freeing up time you’d otherwise spend on non-value-added tasks, and turning that into time you can use toward the value-added tasks.
Chris Hoyt: I agree with you. What’s interesting from my perspective — and Gerry, I’d be interested to hear yours — is that we all labeled it AI, but when interview scheduling first came into our space, it was automation. Very complex “if this, then that” type of automation. That’s when leaders had an “OMG” moment — that AI and automation were really going to have a powerful impact on the business. A lot of those conversations were around, “Why do I have 15 people managing just a scheduling function at my enterprise when I could have one person and a solutions provider?”
Jessica Aguilar: Exactly. I think that’s a gateway into AI for talent acquisition. Where I see the ongoing value is on the recruiter experience as well as the candidate experience. We talk about AI solutions tied to candidate experience, and there’s certainly a positive impact to be had there. But the biggest opportunity I see is taking out some of the noise in the recruiter experience.
I’ve had the opportunity to work with many wonderful talent acquisition professionals, and if I look back through the years at consistent themes, the biggest one is that most of them really love what they do. They love the time they spend with candidates. Their biggest frustration is they don’t get enough of that in their roles — they get pulled into a million other things that aren’t aligned to that core task of identifying great talent, attracting them into the organization, and converting them into a role with the company. So I do see a big value-add in AI capabilities enabling recruiting teams to do more of the things we really want them to be doing.
Gerry Crispin: That’s so on target, Jessica. When you look at this issue, every stakeholder involved in the process is being impacted in one way or another, and every one of them needs to reflect on not only how it impacts themselves, but how it impacts each other — and that balance is still evolving. With your background, do the leaders — not just in your organization, but others you’re familiar with — are they ahead of or behind the folks in the trenches who are playing with the tools and embracing them? Are the leaders actually embracing the tools, or just the concept?
Jessica Aguilar: That varies. I believe in large part they’re embracing the concept. They aren’t as close, typically, to the tools themselves. The piece I don’t want to lose here is the support mechanism between those you’re serving in a recruiting role and the actual role itself. My perspective is that with the evolution of so many of these tools, there’s a bit of a discrepancy in the understanding of what someone in talent acquisition really does. And that’s fine — I’m sure we could apply that to multiple roles.
But what that means is you have individuals who are driving and listening to the radio, hearing ads for AI solutions that say, “Hey, just post your position here and we’ll have 100 candidates for you in the morning.” And then they come in and say, “But why doesn’t it work that way here?” It’s because it’s a more sophisticated process, especially for an organization with a high level of professional roles. If you have to meet multiple criteria to be qualified, the screening process is going to look different than for someone with fewer criteria to meet — then you can simplify things a bit. So there’s a bit of noise in the system. I believe it’s just a disparity in understanding.
Chris Hoyt: You hit on a really good point, and I’ll get even more pointed: I think a lot of those who say “put your job here and you’ll have 100 in the morning” — it’s just bullshit, because a lot of them don’t have any real AI in there. It’s more just search and match, complex “if this, then that,” and they’re slapping an AI sticker on it and calling it a day to sell it. That’s a continual battle across TA, and I think it gets — I don’t want to say more challenging — but more of the same with a fresh coat of paint every time a new shiny bauble comes along.
Jessica Aguilar: Yeah, yeah.
Chris Hoyt: We could do a whole show about all the bullshit in our space. Let’s talk about shifting from automation to actual AI, because I think we’re dancing around it — we want to get to agents, we said we would. When you talk to TA leaders about guardrails around AI, because I think that’s the real thing a lot of organizations are struggling with — can you share what a governance framework looks like in practice for someone who owns the function but isn’t a data scientist? Where do they even begin if they’re looking to include AI in decision-making or recommendations in their process?
Jessica Aguilar: I believe it’s a partnership, especially if you’re in an organization with functions that could support you in a more sophisticated way than you might be able to yourself. Most individuals in talent acquisition aren’t data scientists, they aren’t cyber experts, and candidly they don’t really want to be — but that’s okay, because a lot of us have colleagues in our organization we can lean into.
If there’s a principle of responsible AI that your organization subscribes to, that’s a great starting point — look at that and understand how the commitments you make as an organization line up against what you’d like to do from an AI perspective. Have an assessment against that to understand if it’s synced up, and if there are other things you need to consider to move something forward.
I also strongly recommend that, because in isolation you can only get so far — it really does take connectivity across an organization to move something forward. The worst kinds of surprises happen when you’ve really invested in bringing something to life, only to have it struck down at the eleventh hour because you didn’t take the time to bring the right stakeholders along with you. There’s a reason for that — there are things we do as subject matter experts that support our colleagues, and vice versa. That partnership is critical to successfully identifying and implementing AI and agentic capabilities.
Chris Hoyt: That’s a good transition, because we’re bouncing around the words “AI” and “agentic,” and I think there’s a real difference between an AI tool that recommends and an agent that takes action on its own. We do a monthly lecture where we open it up to our members and alumni, with a speaker from around the world presenting on a topic. We had one do a really great session explaining the difference between AI and agentic AI. If anybody’s listening and would like to watch that, let us know — we’ll get you a copy.
So my question for you: as agents start executing decisions in our hiring workflow, what do you think changes about how a TA leader should think about oversight — about managing the agents?
Jessica Aguilar: We have a very human-in-the-lead focus as part of our responsible AI principles, and that’s something I’d recommend to anyone embedding agentic decisioning. The bigger piece — and it’s the less fun one, I’ll be frank — is auditing yourself, especially at the forefront. When you’re excited and just putting something in, that’s actually the most important time to start self-auditing, to understand if the agent is doing things the way you’d expect, and if there are things happening you wouldn’t have expected.
You can do a lot of that in test, but in a live production environment you learn a lot more, because there’s a lot more data. And the more data you have, the more things can happen. If you have that first step in place, the auditing is actually straightforward, because you already know your aim — what are you trying to accomplish with the agent, what should it be doing, how have you structured it. Do this, this, and this. Don’t do this. Only pull data from this. Only make decisions on this. If it’s doing something outside of that, it’s not too difficult to figure out — and then there’s the remediation aspect.
From an agentic perspective, I find it’s a bit more complicated on the hiring side. From a recruitment perspective, you have to be careful about where you put agentic capabilities. It’s easy to get a little too onerous about where they sit and what they’re responsible for — things that might look good on paper but don’t always serve the candidate experience well. Mapping that is really important.
It’s so easy to build agents now — there’s the piece where you might do it through a third party, and it’s a little easier because they’ve already scoped some of this out for you. But a lot of organizations I’m talking to are also building their own agents, and I understand why, since it’s easy to build them. That ease is something to be mindful of, though, because it creates ease in error too. Second set of eyes, testing, self-auditing — all really important.
At the end of the day, the biggest piece we run into is ensuring we’re aligned from a regulatory perspective, which I wouldn’t want to miss in our conversation, especially for individuals working for multinational companies. It varies — certain US states have different AI regulations, we’re still missing the federal piece, and in the EU there are rather rigorous AI regulations primarily tied to selection and promotion. That hits squarely in the space we play in from a talent acquisition perspective, and I don’t think that’s bad, by the way. I think those regulations are helpful — they support the audit and testing that ensures we’re employing agentic and AI solutions in the right way for the right reasons.
Gerry Crispin: Yeah, the EU AI Act has even been delayed to 2027, just to make sure people can get there from a compliance point of view.
Jessica Aguilar: It’s a lot.
Gerry Crispin: Which suggests a lot of companies are going to significantly change the architecture of their organization — what the organization is composed of, the roles that are now going to be focused on quality control or governance will shift significantly, I’d think.
Jessica Aguilar: The nice thing about the EU Act is there’s been a period of time for feedback and socialization, so it didn’t just drop out of the sky. We’ve had a lot of time to think about what that means in practice. As for how you restructure — yes, quality checks are important, but I’d say that’s more broadly true of AI use in any organization. The way we talk about skills changing, roles changing, the more we use AI, that’s also causing the governance structure of an organization to change.
We have AI champions, and that’s a nice way to embed governance in a way that helps move things forward rather than sounding like a blocker to progress. Those are individuals who consult with you and help you understand, based on your use case, how it aligns from a regulatory and organizational governance perspective — a thought partner in that implementation process who’s there to help you succeed rather than serve as a gate.
Chris Hoyt: You raise an interesting point. We just did the CXR Recruiting Awards — you can check it out, we had 23 submissions, at cxrrecruitingawards.com. You can see what TA teams built. Over the last couple of months we’ve been having conversations with leaders, members and non-members, about how they’re building things from within the TA function.
This raises my eyebrows, because this is super interesting to me. I’ve talked to a handful of leaders who say, “We’re going to build our own ATS, we’re going to build our own CRM using vibe coding,” and I’m thinking, I’m just not there yet — I don’t think we’re there yet. It makes me nervous. There’s a lot of conversation around auditability and accountability. Back in 2023 or 2024, I think it was Marine Layer — they implemented a whole process for handling returns, and the engineer running it quit and went somewhere else, and the whole thing fell apart because there was no structure to how it was built.
So there are a lot of lessons for us to take from the software engineering world. As TA leaders, we’re not scientists, we’re not engineers — for most of us, that’s by choice. But there’s a lot to learn from the structure, execution, and follow-up there, to keep us out of trouble, with some governance and partnerships.
On the flip side, from an accountability standpoint — when an automated part of the process gets something wrong — a good candidate gets screened out, or we get a decision that’s hard to explain — who owns that? I want to double-click on this a little. How do we keep a human genuinely accountable, rather than just pointing at the system and saying, “The AI did that”? Are you having those conversations? Are you seeing others having them already?
Jessica Aguilar: Both. There’s a piece around this — it’s not really an AI topic, it’s been around for a while — the automated disqualification of an applicant. I believe AI has moved that more front and center, and there are more ways for it to happen depending on how you’re structuring your infrastructure, whether it’s your applicant tracking system or a third-party AI bolt-on. But that’s all based on a decision tree that you set as an organization.
If you have a role where the minimum qualification is you either have this or you don’t, and an individual applies and doesn’t have it, that’s a very defensible position. If it’s a gray space, then on your decision tree, I’d recommend that’s not an AI decision — although some might disagree and feel there are sentient capabilities to agents, I’m not there yet, and I don’t believe that exists. I believe that’s the inflection point for a human.
Chris Hoyt: We’re not giving big decisions to a fancy autocomplete box? What do you mean, Jessica?
Jessica Aguilar: I know — don’t ask me where that idea came from, but no, I’m just not comfortable with it.
Chris Hoyt: We did an exercise — we’ve actually done it a couple of times at different meetings — where we task the room, it’s a bit more complex than this, but we task the room with rewriting a process within the function five years from now, and categorizing it as fully AI, fully human, or hybrid.
Announcer: Mm-hmm.
Chris Hoyt: Not a single — and we’ve covered maybe 17 different stages within TA as we’ve done this — not a single room, at this point well over 150 leaders in the space, has had a fully AI solution in five years. It’s always got a human in the loop, which is kind of what you mentioned earlier. I think that’s a powerful statement — you’ve got some people lobbying for it, but at the end of the day, when they’re looking to push something through or be held accountable themselves, there’s always a human in the loop for a lot of that decision-making today.
Jessica Aguilar: And part of that, perhaps, is also putting ourselves in the seat of the people we’re trying to attract to our organizations. If that were us in that seat, being reviewed, we’d certainly want an equitable review from an organization.
Chris Hoyt: Yes.
Jessica Aguilar: I think that’s the other piece to remember — there’s the human-in-the-lead on our side, but there’s also the human side of the individuals we’re trying to bring into our organization and attract to our company.
Chris Hoyt: Do you think this is the same stink, different day? Do you remember when we all got ATSs and could do pre-screen questions, and we had knockouts? When that first hit — you may not have been in the HR or TA space yet.
Jessica Aguilar: I wasn’t, but it resonates, because the idea of knockout questions has been one of consternation for quite some time.
Chris Hoyt: Yeah. As a recruiter, it was gold, though. If you weren’t eligible, or couldn’t make it to the location, or whatever the questions were, bona fide or not, in some cases I didn’t have to look at your resume — I didn’t even have to know your name. So I feel like this is the next evolution of figuring out: are we asking real, bona fide questions, and how are we knocking people out? And are we giving them the opportunity to do more than just a yes/no toggle? Because sometimes it’s nuanced, right? Some of the questions we ask, historically, don’t give people a fair way to answer, because it’s a little more complicated than that, and out they go.
Gerry Crispin: In one way, the knockout question right now is: are you really a person applying for this job, or are you a representative of some kind of person? I have a feeling there are some interesting potential knockout questions in the AI future that are almost immediate and prevent people from completing the application — so we don’t want your expression of interest at this point, because you’re not real.
Chris Hoyt: But if I’m not a real person and I send a robot, I’m probably going to lie on that question. It probably doesn’t matter.
Gerry Crispin: The robot will have to be the one lying — “Oh no, I’m real.”
Jessica Aguilar: Right. I actually love that you said that, Gerry, because I do believe that’s a big conversation right now — with so many “authentic” services out there spraying resumes and application volumes skyrocketing, it’s unfortunate, because it has an adverse effect on the individuals who take the time to learn about your organization and the role. Maybe they’re using AI on their CV — that’s a separate conversation — but they’re genuinely trying to pursue that position, versus someone waking up in the morning to an email saying 100 positions were applied to overnight.
Now, granted, I think that’s kind of AI slop. I bet they’re not getting their money’s worth, because the best way to run that kind of service is to keep people as customers, and the worst way is to lose them because they got a job and canceled their subscription. So the motivation from a business platform perspective seems out of sync with who they’re trying to serve. But there is an absolute impact on whether this was a human who submitted and is really interested in us, or just a bot submission overnight.
Chris Hoyt: Yeah, there’s a spectrum of AI usage, from polishing a resume or slightly inflating responsibilities, all the way to Gerry’s “beep boop, not a real person applying for the job.” That’s the fraud, the laptop farms —
Gerry Crispin: It just seems to me that in the future there should be some kind of pre-engagement, where the representative for the human candidate and the representative for the employer — both agents — agree that this individual, the human, is good enough potentially to be competing for this job, but has to do it intentionally. There has to be a human in the loop on the candidate side, as well as the employer side. I find this a fascinating conundrum that I think we’re seeing the edge of — it’s creating this problem where so many people seem to have applied, but many of them don’t even know that they applied.
Chris Hoyt: Jessica, should we just go back to having people line up around the block to come into our employment office and apply, holding their nice paper resume? I think we just go back to that. Whatever.
Jessica Aguilar: Love it. I do like the agentic idea Gerry’s describing, though. I think that sounds interesting — I’d like to see that workshopped a little, what that actually looks like, because I agree it’s happening today. Today it’s happening: you have candidates using AI to create their CV, to answer pre-screen questions, and you have organizations using AI to screen job seekers. And somewhere in there — it’s actually what keeps me up at night — is that we really have not stopped, in an intentional manner, as practitioners, wherever we sit, to evaluate how this actually needs to evolve.
Because there’s so much noise in the system, and it only continues to rise, and AI hasn’t helped that from a candidate perspective. I read Recruiting Hell just to keep my ego in check, because I know there are a lot of things that are wrong, and it keeps me grounded. Even though some people are just really upset in general, there are things that come out of those Reddit threads that you realize are just the everyday experience people are having, and it’s really not what you want for them.
Chris Hoyt: You should know, I am proudly shadowbanned from Reddit’s Recruiting Hell for occasionally defending processes that do make sense. There’s a lot of crazy in there that shouldn’t be happening, but on occasion, some of it does make sense — and that subreddit is just something.
Let me ask you, Jessica — if a TA leader is listening right now, and they know they need to move on this but feel behind, feel like they’re running late on this — is there a responsible first step you’d recommend? What separates the teams getting this right from the ones that are going to regret how they rolled it out?
Jessica Aguilar: I do believe there’s a responsible first step. There are so many shiny objects out there, and so many people coming at TA leaders saying, “Why aren’t you using this? Why aren’t you doing this? Why don’t we have this?” The process I feel strongly is the right one is to take a step back and really evaluate your system, your process — talk to people, understand where the pain points really are. And when I say talk to people, that’s your hiring leaders, your candidates — which most organizations get feedback from through NPS surveys or similar — and also your team, the people actually using your systems. Determine what would make the biggest impact, what are some quick wins, and what requirements you’d need in order to do that.
What I’ve observed is a lot of organizations feel pressure to implement something, and then that something fails, because it was never aligned to what they really need to show value to their stakeholders. Culture is big, even in an AI implementation — you don’t want to lose your company DNA just because you’ve taken on new technology. You want it to complement who you are, to accelerate your ability to do the things you’re capable of doing, and to attract the right talent.
So I’d tell them to really block out that noise of “do, do, do right this moment,” and be surgical about what you actually need now — and what kind of process re-engineering actually works here. You can’t adopt AI without process re-engineering. If bringing in something could be great, but you have a team that would never change out of the gate, give them a gateway — something that helps win them over to adoption, so that the next piece you put in front of them, they’re much more likely to pick it up and use it.
Chris Hoyt: Well said. What I love most about that, Jessica, is as we’ve been talking through this, not once have you said you’re willing to forfeit candidate experience, or forfeit letting candidates feel heard, or fully express how they believe they’re qualified, or get feedback. Not once have we said we’re throwing that out in the name of automation, speeding things up, and doing more with less. It’s a balancing act, and it’s not easy for an enterprise organization.
Jessica Aguilar: No, it’s not.
Chris Hoyt: Well, Jessica, we’re going to let you off the hook, but you’ve got one more question — we ask this of all our guests before we stop the show. If you were going to write a book on this topic, what would the title be?
Jessica Aguilar: Fall Hard, Get Up Fast. I don’t know.
Chris Hoyt: There’s nothing wrong with that. Fix your teeth and keep moving.
Jessica Aguilar: I’m obviously not vying for an author title anytime soon, either.
Chris Hoyt: Oh my god, that’s so funny. Well, Fall Hard, Get Up Fast, I love it. Who would you give the first signed copy to?
Jessica Aguilar: My husband.
Chris Hoyt: Oh, that’s lovely.
Jessica Aguilar: Yes — the person who puts up with me every single day. He deserves something.
Chris Hoyt: I love it. Great. Well, Jessica, I know you’re super busy — thank you so much for dialing in all the way from around the world, giving us some of your time today. We really appreciate it.
Jessica Aguilar: I appreciate the invitation. Thank you.
Chris Hoyt: Always happy to connect. For everybody who hung in there with us — whether you’re still on the treadmill or doing whatever you do — just a quick reminder, cxr.works/podcast. You can check out other episodes, or listen to this one again. And again, I’ll remind you: cxrrecruitingawards.com, if you want to check out the submissions that were created by TA teams — teams actually built these AI solutions within their workflow, and some of them are pretty darn impressive, so I’d encourage you to check those out.
And I guess that’s a wrap. Gerry, when will we see everybody again?
Gerry Crispin: Soon.
Chris Hoyt: Yeah, next week. All right, everybody, have a great week.
Announcer: Thanks for listening to the Recruiting Community Podcast, where talent acquisition leaders connect, learn, and grow together. Be sure to visit cxr.works/podcast to explore past episodes, see what’s coming up next, and find out how you can join the conversation. Whether you’ve got insights to share or want to be a guest on the show, we’d love to hear from you. If you’re interested in learning more about becoming a member of the CXR community, visit us at www.cxr.works. We’ll catch you in the next episode.