Hello, everyone, and welcome to today’s session. My name is Conor Volpe. I’m a Product Marketer from Eightfold, and today, we’re going to discuss how AI and TA come together.
I’m sure I don’t have to tell many of you that there are massive implications for AI across all industries — but particularly for TA. There’s been a lot of talk, ideas, suggestions, and technology that’s come out over the past few years.
I imagine for many people listening right now, maybe you’re somewhere between excited about the possibilities for AI, maybe you’ve got a bit of AI fatigue because you’ve been hearing about it so frequently for a while, and maybe you’re feeling a little overwhelmed because it seems to be coming on so fast and it’s hard to understand all the different ways you could use it.
Point is, I don’t blame you one bit if you find yourself somewhere between all three of those.
So for today’s session, we’re going to use this as a bit of a level-set. Yes, AI is everywhere. And yes, it can be overwhelming, fatiguing, exciting. So let’s take a step back and focus on the basics.
At a very high level, how does it work? What are the applications for AI in Talent Acquisition? And how does it help professionals like yourselves?
I’m glad we get a chance to do this because these are the kinds of topics we at Eightfold help our customers with every day — figuring out how to best leverage AI for the benefit of their talent and their talent processes, to ultimately achieve their talent goals.
So for today’s session, we’ll cover a few areas. First, we’ll talk about the state of AI and a bit about why there’s that kind of triangle between excitement, fatigue, and overwhelm that I mentioned before. We’ll have a little bit of fun with that.
Then we’ll apply AI to TA, thinking broadly about how the two can work together. And finally, we’ll get into some use cases for AI and TA so you can see the ways that some companies around the world are actually using AI to their benefit today.
Before we get into all this, I mentioned that triangle, right? Overwhelm, fatigue, and excitement. And many, many people I’ve talked to are dealing with that.
AI is everywhere. It’s really hard to tell sometimes what it is or why I should care. But at the same time, there’s a lot of urgency and immediacy that companies are dealing with and talking about when it comes to adopting AI.
So it’s clear — it’s really important. Everybody’s moving in this direction. So what do we do with it?
Oh, let’s be clear. AI isn’t new. But man, has it been really popular over the past few years. To the point that news about AI has quickly become front-page international news — as much as we have front pages nowadays.
The point is, it’s become global. It’s not industry news. It’s not tech news. It’s not only news for people who follow AI. These have been big events.
And to me, this change — this AI becoming global news — that happened with the introduction of ChatGPT in 2022. Again, this was not the first piece of AI, but it became really important because we had it in our hands. We could see generative AI creating answers for us, giving us essays.
We heard about it passing tests, and it was really tangible. We could literally see it writing out responses. Maybe some of you have kids who are using this as a work aid today.
So we started to realize the benefit, and companies saw all this potential coming out of technology like ChatGPT. And it created this whole new level of excitement around AI.
Not too long after the introduction of ChatGPT — and partially because of the excitement around it — the EU put out sweeping AI regulations, the likes of which the industry hadn’t seen before. These covered everything from clearly labeling deepfakes, to regulations for certain AI applications in high-risk areas like hiring and education, which put more scrutiny and transparency requirements not only on the companies that use these AI applications, but the companies that make them.
They also just outright banned certain AI practices, like social scoring. In large part, as I said, this was a reaction to the excitement and potential behind ChatGPT. Some thought it might temper that enthusiasm — but it didn’t. Far from it.
In fact, OpenAI — the parent company behind ChatGPT — raised a $6 billion funding round, putting their valuation at $157 billion. Truly a mindboggling number.
And they aren’t the only ones doing this, right? There’s Claude. There’s Gemini. Meta’s got LLaMA. There are all kinds of models out there.
The proliferation of this technology likely hasn’t even peaked yet. And all that happened from 2022 to 2024. I know 2025 is barely underway at this point, but we’ve already seen big AI news stories this year.
There’s Stargate — a joint announcement where private companies are banding together with the U.S. government to put $500 billion of AI infrastructure into place. Or there was the DeepSeek announcement not that long ago that took a trillion dollars off the U.S. stock market.
For those who aren’t super familiar with it, DeepSeek is a Chinese company that created a ChatGPT-like competitor. That by itself isn’t necessarily big news, but they did it for a fraction of the cost — and the ripple effects showed up in the market.
All of this is to say: AI is really popular. I know it, you know it, we all know it. But just how popular is it?
To figure that out, we’re going to have a little bit of fun — introducing the international debut of the game called Guess What’s Trending. This is a game show where I give you two choices. One is an AI-related topic, and one is just another topic. You’re going to say either in the chat or just to yourself — write it down on the honor system — which you think is more popular.
The proxy we’re using for popularity is Google search results. Is it perfect? Probably not. But it gives us some insight into what people are looking for, trying to understand, learn more about, or get access to.
So this is over the last 12 months, in the U.S. only. But think about it — over the last year, which of these two topics has been more popular according to Google?
Our first part of Guess What’s Trending is: “Large language model” or “Coca-Cola.”
Obviously, large language models — this is what ChatGPT, LLaMA, and all these other systems are built on. Coca-Cola, on the other hand, is one of the most recognizable brands in the world.
So again — in the chat or to yourself — which of these do you think people searched for more on Google over the last year: large language model or Coca-Cola?
The answer, according to Google, is Coca-Cola. You can look at the bar chart on the left to see which has massed more interest over the last 12 months. You can see Coca-Cola has maintained popularity for the most part over large language model. But just recently, large language models had a spike — probably because of the DeepSeek announcement.
So is the AI topic more popular than Coca-Cola? Maybe not. But Coca-Cola is probably the most recognizable beverage brand in the world.
Next up for Guess What’s Trending, we have: ChatGPT or taxes.
Which do you think has been more searched? Type it out if you’ve got a guess.
If you said ChatGPT, you’d be right — with an exception for tax season, where taxes were briefly more popular. But for most of the year, more people were searching for ChatGPT.
And finally, we saved the real test for last. I ask for some preemptive forgiveness from any Swifties out there — I hope I don’t offend anyone.
Which was more popular over the last year: artificial intelligence or Taylor Swift?
Again, in the chat or to yourself — which one: Taylor Swift or artificial intelligence?
Feels like we need a drumroll sound effect or something for this reveal…
If you guessed artificial intelligence, you’re right. Over the course of the last year, it was more popular according to Google than probably the most famous or well-known musical artist in the world.
So where does all this stand? Well, if you wanted to know how popular AI topics are — they’re somewhere above taxes and Taylor Swift, and below Coca-Cola.
Let’s step out of game show mode for a moment. I hope that was a little bit fun and helped you see why you’ve heard so much about AI and why it’s kind of everywhere — because it really is.
But the popularity of AI makes this statistic even more interesting. By the way, this is from our own talent survey that Eightfold does with HR practitioners and leaders. 96% of companies are either planning to use, or already do use, AI tools and technology.
Which means almost everyone is in on AI — whether they’re outright using it today, experimenting with it, or know they will use it in the near future.
So it’s not just you and me plugging emails into ChatGPT to optimize them or whatever that might be. Enterprises are looking to get the most out of their AI too.
And I imagine many of you on the line are doing this right now — whether it’s you being proactive and getting ahead of the game, or someone is asking you to bring it into your workflows today.
But almost in the same breath — in that same survey — most are worried about going too fast. Change has always been constant, that’s a given. But right now it can feel especially constant — and perhaps even disruptive — for the people we talk to.
Because while AI can be very useful, it’s also forcing companies to perhaps change in ways they weren’t expecting to. And that means adopting AI at scale — faster than many would have thought they’d adopt a new technology.
So with that in mind, our goals for today are about keeping it simple. I hope you walk away understanding just a few ways that AI and TA work together today.
This is not the be-all-end-all. You’re not going to walk away with a master’s-level thesis on AI. But I hope that by providing some very tangible examples, you get a little bit more comfortable — and ideally, a bit excited.
Because the implications for AI and TA are really, really interesting. It’s exciting to see the kind of impact this can have on talent.
I hope this helps reduce some of that fatigue and overwhelm — and brings up a little bit of that excitement.
Before we dive into use cases, though, I want to cover a quick primer on what to consider when you’re evaluating AI.
The first thing to consider when you’re evaluating any kind of AI technology that you’re bringing into your company — and there are so many — is the type of data that it has access to. Or really, what fuels it. This is what powers the AI.
You’re all well aware of the phrase “garbage in, garbage out.” That has never been more relevant than with AI.
The data that the AI uses to make its insights — to give you insight — should come from three different places. It should be a mix.
The first is your own data. It should use your internal data so it knows your company, how it works, who succeeds, how long requisitions get filled, how long employees typically stay, what the satisfaction levels of your recruiters are — all those things.
They help the AI learn from you. Not every company in the same industry is the same. Not every company of the same size is the same. Not every company in the same country is the same. So it needs to understand the intricacies of your organization in order to deliver useful insights.
At the same time, the AI should have access to external data. If the only data it’s using to give you insight is within your four walls, it limits the quality of what it can do. Bringing in external data helps prevent bias and avoid a self-fulfilling prophecy.
Let me give you an example. Let’s say traditionally, your company has hired entry-level engineers from local colleges. Maybe that was a size thing — but now your company is expanding. You have other offices, and you’re looking to hire engineers in those locations too.
Well, perhaps the schools around your HQ have traditionally provided you with mechanical or electrical engineers, so that skill set has been associated with your entry-level hires. Same with location.
Now that you’re expanding, maybe you want different skill sets. Maybe you’re open to different geographies. But the AI might associate “success” only with the engineers you’ve hired in the past — their location, or their exact background — and that can lock you in.
Access to external data allows the AI to understand that successful entry-level engineers don’t have to come from a single school or region, and they don’t all have to have the exact same skills.
The third leg of this data stool — once you’ve brought in your own data and external data — is user interaction. This is the handshake between the AI and the human.
User interaction should guide the AI. It should be able to see which insights or recommendations are actually making an impact.
This can be implicit — the AI paying attention to where users are seeing success. And it can be explicit — users giving direct feedback or making changes that guide the AI.
Let’s say recruiters now say, “Actually, we want a specific skill set.” The AI can now help optimize for that. But even if they don’t say that explicitly, if recruiters start hiring more candidates with those traits and those candidates succeed, the AI should pick up on that trend.
Now that’s the data piece — obviously very important. Let’s also consider the AI itself, or the engine. What is actually using the fuel?
This is where we could go really deep, but for today, let’s focus on one core question: What is the AI built to do? Why was it built? What is its purpose?
I say that because all AI is built around a hypothesis. And if you understand what the AI was designed to do, you can figure out if that purpose aligns with how you want to use it.
Here’s a lighthearted example. This past summer, my wife and I were invited to one of our best friends’ weddings. Anyone who’s planned a wedding knows your to-do list is about 111 items long, and you don’t always get to everything.
They asked us to help come up with a punny, funny hashtag they could use on Instagram. We said, “No problem!” I turned to my trusted friend ChatGPT to help get us started and provide inspiration. We gave it the bride and groom’s last names.
While ChatGPT didn’t get it wrong, it didn’t really get it right either. It gave us hashtags like “LastNameLastNameForever” or “LastNameLastNameGetHitched.” They were technically correct. They just weren’t funny.
At the time, ChatGPT wasn’t built to be humorous. I think it’s gotten better since then, but the point is: the AI wasn’t designed for that task.
Same thing applies in TA. If you use an AI built to optimize workflows and ask it to evaluate a candidate, it might not give you the best result.
So always consider why the AI was built.
Finally, we have what really matters — the insights. What are you actually getting out of it? What value does it provide?
And that brings us to our problem set.
I’m sure I don’t have to tell any of you that you’ve run into these problems before — or are running into them today — or will in the future. These are very common TA problems that AI is helping with.
Is AI going to solve these altogether? No. But it’s going to help TA organizations make big strides.
And hopefully, by going through these and showing you some practical examples, we’ll help you understand a bit more about how AI applies to TA today.
So with that, let’s dive into our first use case.
I want to start from the beginning — actually building a req.
There are so many factors that go into building that req and getting the position started. But one that we’ve seen can be nice to have for some companies — or even aspirational — is planning for potential.
How do we make sure that the people we hire are not only going to be able to do the job today, but can grow with the company?
Do they have the skills the company needs? Do they have the skills the company has too much of? Are there gaps?
The right AI can show recruiters those skill-based insights from the very start. Recruiters can build the req with these insights from the jump.
Ideally, the recruiter can see which skills their company already has, which it may lack, and which ones are trending up or down in the market. That way they can prioritize what to look for — not just based on the job today, but on long-term potential.
They can even benchmark against the market and the industry.
This means conversations around skills, potential, and the dynamic nature of talent can happen from the very beginning — as you’re writing the job req.
Once we’ve built the req and posted it, we’re looking for candidates. But traditionally, that starts with a blank pipeline.
As we mentioned before, AI should have access to external or public data — and it should use that data in its insights. With public data, you can start to update old candidate records so they’re usable again.
That means all those silver medalists and alumni sitting in your ATS or HRIS — those profiles might be years out of date. But they don’t have to be.
Sure, you could go manually look up their LinkedIn profile and update what they’ve done over the last few years. But AI can do that at scale.
It can help you find people you already know — who maybe weren’t the right fit before — and see if they’re the right fit today. It can enrich their profiles so you can assess them for the current role.
That includes previous applicants, silver medalists — essentially anyone sitting idle in your system of record. Now they can be re-engaged for this new role.
So instead of starting from scratch, instead of starting from zero, you’re now starting from a warm pool of potential fits — with updated, usable information.
And you should also be able to see how much experience they have, what’s changed, and why they may be worth a second look.
Now let’s go to the opposite problem. Maybe you’ve got too many applicants or candidates, and it’s hard to understand who might be a good fit.
Again — this is a scale problem. AI can help with that.
But even better, let’s go back to the data question we discussed earlier. AI understands your company through your data. It understands the types of candidates who have had success at your company.
So if you’re hiring for a project manager, there are lots of different types — different industries, different skills. Which of those traits have led to success at your company?
The AI should not only be able to determine that — but show it to you through explainable AI.
So even if a candidate doesn’t look like a perfect fit on paper, maybe their skills do align. Maybe the industry they came from maps well to yours. Maybe their job titles translate more than you’d expect.
Whatever the reason is, the match score the AI provides should come with an explanation.
That’s key. This match score isn’t ruling people out — it’s ranking them so that recruiters have more data to help make decisions.
AI is not making the hiring decision. This is still a handshake moment — the recruiter stays in the loop.
Explainable matching helps recruiters understand why a candidate is being recommended and enables them to make more strategic choices as they move candidates forward.
Let’s talk about the career site — where candidates apply.
I’m sure you’ve all done this before. You’ve created the pipelines, posted the reqs on job boards, maybe even done some outreach.
But then a candidate lands on your site and sees a wall of jobs. They have to apply filters, dig through listings, and guess which ones might be a fit.
They have to scan job descriptions, requirements, and try to decide if they’re qualified.
That’s a leap of faith on the candidate’s part — and it’s where companies often lose applicants.
Because candidates don’t always know if your definition of “account executive” matches their previous experience. It’s hard to tell.
AI can help flip that process.
Instead of forcing candidates to search and filter, AI can search for them.
A candidate can upload their resume. The AI parses it, understands their experience, and matches it to open roles — just like it would match a candidate on the recruiter side.
It can then surface jobs the candidate is a fit for and show why. Explainable AI gives the candidate more confidence and helps them apply to roles that make sense for their background.
Let’s talk about sourcing.
We all know that maintaining talent pools takes a lot of work. When done well, it can be very demanding.
But if AI can match candidates to jobs, there’s no reason it can’t also match candidates to content or talent pools.
Let’s say you want to create a pool of silver medalists in a certain geography, with 5–10 years of experience in machine learning. The AI can take that input, search your existing talent, and create the pool for you.
Even better, the AI can parse content — like an article, video, or blog post — and identify the key topics or themes. It can then match those to the right audience in your talent pool.
That means more relevant content going to the right people — and less manual work for sourcers.
It takes some of the burden off, so sourcers can focus on building relationships, not segmenting lists.
Finally, let’s talk about recruiter time.
You all know how many repetitive, manual tasks recruiters are burdened with every day. And HR tech vendors know this too.
That’s why many of them are introducing GenAI features. And the first place many companies test these features is in recruiter workflows.
You’re probably familiar with automation tools — “If this, then that.” They help move things through the cycle faster.
GenAI can take that to the next level: orchestration.
That means the AI can complete tasks on behalf of a recruiter — things like generating a job description or drafting an email.
But it doesn’t stop there. It can also interact with the recruiter to fine-tune those actions. This is another handshake moment.
AI should never act on its own. The recruiter should always stay in the loop.
With orchestration, the AI becomes a co-pilot — helping the recruiter do things like launch a new role, build a talent pool, or invite candidates to apply.
These are multi-step tasks. AI can help streamline them, but the human is always the decision-maker.
These tools are huge productivity boosters and time-savers for recruiters.
So that was a lot. We covered a lot of ground.
But again — our goal was to give you a glimpse into how AI can help TA, from writing the req all the way to selecting the right candidate.
Hopefully, we helped you get a little bit excited about AI, and showed you how companies are using it today in real talent workflows.
Again, this is the kind of work we at Eightfold are passionate about.
Our goal is to help people apply AI to their talent — and this is what we do every day.
At the center of it all is our Talent Intelligence Platform, powered by one of the largest talent datasets in the world. We help customers transform their talent practices through this platform.
Some of this is individual guidance — for employees on their career development, career pathing, mentorship, and course selection. And some of it is recruiter and candidate guidance — helping people find the right opportunities based on skills and potential.
It also includes organizational guidance — helping companies source, screen, plan, and redeploy talent with greater confidence and speed.
All of this brings us closer to our mission as a company: to help find the right career for everyone in the world.
So, if you have any questions about anything we presented here today, please drop them in the chat. Let us know — we’d love to hear from you.
Stop by our booth! We’ve got the latest research our team has sourced and created — including the survey I mentioned earlier.
We’ve also got demo videos and a whole bunch of assets and resources for you to check out. Come learn more about us and what’s happening in the world of AI.
But really — just come chat. We’d love to hear from you.
So with that, thank you for your time today. I really appreciate it.
Enjoy the rest of the event.
Take care.
