Get my free whitepaper “10 Ways AI Will Transform Talent Acquisition” – Download Now
Trend Spotting – Find out how my digital course will help you shape the future of talent acquisition in your organization – Click Here
It is becoming clear that job seekers’ use of AI tools in the application process has profound implications for talent acquisition. Many employers report a significant increase in application volume, and there is a potential technology arms race as employers and job boards attempt to use AI to identify AI.
While this may help in the short term, it won’t work in the long term, and is job seeker AI use even a bad thing?
My guest this week is expert job board consultant Alexander Chukovski. Alexander has been doing a deep dive into how job seekers use AI and its implications for recruiting. He has valuable insights about job seeker verification and its potential to create win-win situations for candidates, job boards, and employers.
In the interview, we discuss:
• The implications of a new generation of AI-savvy job seekers
• The different ways job seekers are using AI include
• ATS integration
• The drawbacks and risks for candidates using AI tools
• A pointless tech “arms race.”
• Why is it difficult to spot the use of AI in CVs and applications?
• How should the industry respond?
• The rise of verification technology
• What can be verified now and what might be possible in the future
• The importance of job seekers owning their data
• What does the future look like?
Follow this podcast on Apple Podcasts.
Matt: Support for this podcast comes from CV Wallet and its verification platform for recruiters. In a world of job seeker adoption of AI and too many unsuitable applications, CV Wallet puts trust at the heart of your hiring by verifying your applicants at the start of the recruitment process, saving recruiters time and money by getting straight to the right candidates for your jobs. Its bluetick verifications are very easy to set up. Work globally are completely bespoke and can be set at an organizational or job level depending on recruiter requirements. Typical verifications include identity checks, right to work, qualifications, licenses and skills enabling job seekers to fast track their applications while securely saving their verifications for future use in their free CV Wallet app. It’s free to sign up and with no integrations required, fits seamlessly within your existing application process. Get started today at cvwallet.com.
[Recruiting Future theme]
Matt: Hi there. Welcome to Episode 612 of Recruiting Future with me, Matt Alder. It’s becoming very clear that jobseekers use of AI tools in the application process has profound implications for talent acquisition. Many employers are reporting a significant increase in application volume, and there’s the potential for a technology arms race as employers and job boards attempt to use AI to identify AI. While this might help in the short term, it won’t work in the long term. And is job seeker use of AI even a bad thing? My guest this week is expert job board consultant Alexander Chukovski. Alexander has been doing a deep dive into how job seekers are actually using AI and the implications it has for recruiting. He has some very valuable insights to share about job seeker verification and its potential to create win-win situations for candidates, job boards, and employers.
Matt: Hi Alexander, and welcome to the podcast.
Alexander: Hi Matt, thank you for having me.
Matt: It’s an absolute pleasure to have you on the show. Please could you introduce yourself and tell everyone what you do?
Alexander: Yes, sure. I live in Munich, Germany. I have been in the talent acquisition space and HR tech for almost 15 years now. Worked in the past with companies like Experteer and JobSync. Today, I mostly advise job boards and aggregators and other HR tech vendors on all kinds of topics like SEO, AI automation, job scraping, parsing, matching, pretty much anything you can think of that a job board might be interested in. And I run also three niche job boards in crypto and in AI. I love to write about the industry. Huge fan of the podcast. Thank you again for having me.
Matt: Thank you. And as I say it’s brilliant to have you on the show. Lots of things that we’re going to talk about, but I suppose just to kind of set the scene a little bit, I mean, tell us about what’s going on in the market at the moment? What are the major challenges that you’re seeing?
Alexander: Yeah, sure. So, I think one thing that is very interesting is Gen Z entering the workforce and this is not something new, but now they are in high numbers. I think they just recently surpassed baby boomers as a major workforce and they’re very technically savvy. So, they’re very AI affine from the beginning. They consume a lot of video content, they learned all kinds of tricks on how to use GPT and AI in their jobs. And I think we’re starting to see this impacting our market. I think we’ll talk about this a little bit later. But when you pair this with the white-collar professionals who lost their jobs in the 2022, 2023 period in the US specifically, also very tech savvy people, really working well with technology and understanding the fundamentals of how automation and AI work. I think all these combinations are setting up our industry for a very interesting future development.
Matt: Yeah. No, absolutely. I mean, let’s dive into that. What is it that you think this is going to drive in terms of change in the industry and how’s that going to pan out over the next 18 months or so?
Alexander: Yeah. So on one hand, of course, you have recruiting, which is prone to automation. We hear a lot about automations using AI and AI has been more or less a bit commoditized. So today, you can just consume AI responses from an API from OpenAI with just a few clicks. And this is of course enabling all kinds of industry specific use cases to be automated. Our industry runs a lot on text, job descriptions are text, people exchange emails. And the AI models that we use today are great at working with text. They don’t understand it, but they are great at manipulating it, extracting information out of text. So, we see a lot of the big players going into the direction of building job description generators. I’m hearing matching again, that’s great because matching has been a tough problem to solve. But there’s also the other side in the market and for the first time ever, I think that we are seeing job seekers going into the AI tools in their application cycles. And this is something that we have not seen in the past. This is completely new. So, this automation goes both ways, right?
Matt: No, absolutely. And what are the implications of that? What are we seeing now? How is that likely to escalate?
Alexander: Yeah, so it’s been fascinating. I recently learned about this trend around tools that can help job seekers apply to jobs automatically. And it’s shaking up the industry a bit because this AI power automation is the first time that it’s actually rolling out to job seekers. And they are actively utilizing it in their job search. It’s been very fascinating. I even had to create a TikTok account, because a lot of these tools are firstly promoted there. We don’t see them in our usual advertising channels. Very interesting things being displayed on a daily basis. And the first time that the industry is seeing how these tools actually all come together and what it means for job seekers. So, put yourself in a position of a young person so you’re technically savvy, you’re looking for a job or you might have been recently let off from your job. So, you want to get a job quickly because you’re digitally native, you learn about these tools and then you ask yourself two questions, “If I tailor my resume to job description, is this going to improve my chances of getting a job? What do you think?” Generally, “Yes, of course.”
And then if I’m able to send a lot more resumes that are perfectly tailored to a job description. Is this also going to increase my chances of getting a job? Absolutely. And when people think like this, we see that these tools actually have a place in the job seekers daily process. And there was a, I think we all read the wired article about a software engineer who applied to 5000 jobs using a tool called LazyApply. He landed 20 interviews. So, compared to his previous success rate of doing 300 applications and getting, again, 20 interviews, but having to do all this hard work, “What do you think he’s going to use in the future?”
Matt: Of course, the LazyApply, which is interesting title for it as well.
Alexander: Yeah, absolutely. And we have 80 tools like this today. I tracked them all. I’m really fascinated. There are different categories. They’re not all looking into creating as many applications as they possibly can. Some are very helpful because they allow you to track your applications, where you apply too, where did you send an application, where did you get a response? Others are just a plugin that runs on top of your browser and they just fill out the apply forms for jobs that you have picked out on certain ATSs. And others will automatically pick jobs for you because they scrape ATS career pages and they have found ways to fill out the application forms using RPA, others will go on LinkedIn and Indeed and again do searches for you. And once one of these tools onboards a job board. And job boards nowadays have also integration to ATSs, essentially, they become meta integrated into these, so you are then able to send even more applications at scale.
Matt: Wow, that’s really interesting. And the irony that as an industry, we spent decades trying to perfect integration. [laughs] The job seekers are just doing it for us straight away. Obviously, I want to really talk about the implications for the industry and what job boards are doing and how this moves forward. But before we do, I mean, what are the drawbacks and risks for the candidates who are using these kinds of systems? Because as you say, “It would make perfect sense for them to do that.”
Alexander: No one actually talks about risks. And for someone who has worked in the past into the whole ATS integration life cycle, when you use, for example, RPA to fill out an apply form, some of the ATSs out there are actually able to detect these applications. So, they’re going to display a success message as if the application was submitted, but then it’s going to be discarded in the backend because there is some kind of suspicious behavior happening. What can also happen is that some of these tools are not going to fill out the full apply form. They’re just going to focus on the required screening questions and they will leave the other blanks. And as you know, depending on how you configure your ATS, again, this application might be discarded because they just assume that if you did not make the effort to fill out the full application form, then you’re probably not a qualified candidate at all. So that’s pretty bad for the job seekers because their resume might not even land on the recruiter’s desk at all.
Matt: No, absolutely. And that’s interesting, as you say, that’s not something that gets discussed in these TikTok conversations. Presuming though, those things aside, there’s still a vast quantity of applications that are coming through. And again, technology might improve and that might increase. What does the industry do about it? What are you seeing job boards and recruiters doing already to deal with this issue?
Alexander: Well, to be honest, the only thing that I have seen is that some job boards are rolling out CAPTCHA verifications from Cloudflare for example. So basically, it’s behavior-based system that says, “Okay, you’re most likely a bot, so I’m going to show you a CAPTCHA, and if you cannot complete it, you cannot apply.” But of course, these tools are not perfect. So, what can happen is that occasionally you won’t be able to apply even though you’re not a bot, just because some model was triggered and assumed that your behavior was suspicious. So that’s pretty bad because it adds friction. And as you are perfectly aware of, our industry has been spending so much time on removing the friction in the application process. So, we are going back to that. The other thing that I have seen is that recruiters often write in the job description saying that they prefer more genuine, authentic applications. So, they would not encourage you to use GPT. But come on, like people just ignore it, right.
[laughter]
Matt: Yeah, absolutely. And also, it just seems like the start of some kind of technological arms race that doesn’t actually ultimately solve the problem, does it?
Alexander: No. I mean, just adding more friction, we know who is safe. Workday on the other hand, they’re pretty safe because it’s so hard to apply there that probably most tools will fail. So, they have found a way-
[laughter]
Alexander: -to be on the safe side.
Matt: That’s not really the blueprint that we want for the rest of the industry.
Alexander: Definitely, no.
Matt: So, what is the solution here? I mean, there’s obviously some radical changes in terms of how we think about recruitment and how we think about applications. What’s the kind of the strategic, long-term way to move forward and really make recruitment better for everyone and make it fairer and use this as a real sort of positive?
Alexander: Yeah, sure. Let’s look at the options. So, ATS platforms today do have some kind of defense against this prey and prey approach, but its mostly keyword based. So, if every resume is [unintelligible 00:14:12] and it’s perfect, then this filter is not going to work anymore. So, people say that they can detect AI-generated text, but this is not really so trivial, especially if you’re dealing with short text in resumes, for example, where you have just a few bullet points if they are generated with AI. Technically speaking, it’s very hard to detect if this is GPT text because even small text changes are going to trigger most current AI detectors, useless because they won’t be able to recognize that this is AI-generated text. And we actually need longer text for these to work pretty well. So, then we could do maybe skill-based hiring. But again, in a post GPT world and on TikTok, we actually see the new tools that are hitting the market. They actually run in parallel on your interview. So, they capture what the recruiter is asking you and they are writing the answers in real time. That’s also fascinating.
So all these Q assessments, like online interviews become also a little bit flawed. So, recruiters get a lot more applications, perfectly tailored, hard to filter them. So, you either have to invest a lot more time in the initial screening and the interviews, which also becomes a bit hard or there is another direction. And I see two possible directions to go. On the one hand, trying to detect these tools, which we talked about, and we see that people are already trying to do. But I think that this is going to be really hard and it becomes like a cat and mouse game or maybe similar to what CV Wallet is building. There will be platforms that are going to merge, and on these platforms, people can verify certain parts of their profile, even identity. Along the verification of the identity is going to be a large step that is going to reduce the amount of unqualified applications, because if you have to do the work to identify yourself, then you’re probably not going to attract people that are interested in using some kind of automated tools. Because they have to do extra work to do this.
So, I think these platforms are going to merge. They will be market specific, they will be niche specific, because different aspects of a user’s profile have to be verified. And as long as these platforms allow people to keep their information on their own mobile phones or computers, wherever they use to apply, people will have an incentive to come back because you’ve already done the work. So as long as I get relevant jobs suggested, then I’m always going to come back to the same platform.
Matt: Yeah, I think that’s interesting. And I know that you mentioned CV Wallet there, who are doing some really interesting work in this area. I mean, I suppose to dive in a little bit deeper into this. We talk about verifying people and verifying their identities and things like that. What’s possible at the moment? What kind of things could be verified by technology? And how might that develop over the next few years?
Alexander: Yeah, that’s a good question. Because if we don’t have the tools to verify profiles, then this whole concept becomes just a concept. So, what we could do today, pretty easy, is anything around identity. There are APIs for this. It’s highly automated, it’s very low risk and it’s been actually used across all kinds of different industries. So that’s fairly easy to do today. You already see LinkedIn doing this and I think it’s just a matter of time for them to start prioritizing applicants based on whether or not they have completed the LinkedIn verification. So that’s definitely possible today, and it is a major step. You can verify, for example, work permits using GPT and the image processing functionality today. There are also APIs from state-based agencies, like in different countries where you can, for example, verify a driving license.
There are even some basic infrastructures out there that can be used to verify degrees and other credentials of this form. But of course, there’s also a lot of work to be done and it’s very market specific, so it really depends on which country you go to. Like in Germany, for example, we have this problem with applicants from foreign countries that don’t speak Germany, but because of GPT, it comes over if they are very good at speaking German. So, for example, one way to verify if they know the language is to look into the degrees that they have. Because if you have finished a German high school, then of course you can speak German.
Matt: Now that makes sense, Is that sort of a probability? It’s like 60% likely they speak German as well as they say, or 10% likely or whatever. Is that sort of working on that kind of principle?
Alexander: Yeah, yeah. But you can just, let’s see, let’s stay at work permits, very important topic. Because if someone applies for a job and they don’t have a work permit for this country, then it’s an unqualified application. It doesn’t matter how good the candidate is. And this is something that can be verified fairly easy because you normally have a passport in your stamp if you’re a foreigner and you just have to make a photo of it, then we send it to an API, and this API says, “Yes, this person is authorized to work and this one is not.” It’s very simple.
Matt: I suppose the really interesting principle here is the candidate owns that verification. It’s like once it’s been done, once they have that, and they can use that to apply for other jobs as well.
Alexander: Yeah. I think that’s the important part. That as you complete certain steps of the verification, that you actually have this data and you can reuse it in the application process for your next job and you’re actually the one that owns it. So, it’s not just standing somewhere on some clouds and everyone has access to it. It’s good if it’s really stored on your mobile device. It’s only utilized whenever you want to apply. This is what I guess CV Wallet was originally based around. And I think it’s a good path. So, if everyone else follows this, great for applicants and their privacy.
Matt: Yeah. No, absolutely. And the overall candidate experience as well, because it will make things quicker. As a final question, it’s obviously a truly disruptive time of the minute. And I know we’ve been saying that for years, but it seems like that’s now the norm. And there’s a kind of real sense that anything can happen. So, I’m just interested in your view about what the sort of medium, long-term future looks like. If we were having this conversation in five years’ time, how might we be talking about recruiting then?
Alexander: Yeah, that’s a very, very interesting question. So, I thought a lot about it at a conference last week in Barcelona, very big HR tech conference. And I saw all the big players actually going back to building their own candidate databases. It seems that a lot of technology is being rolled out into reactivating these candidate databases, and I think that what they’re not announcing yet is the verification steps that they’re going to add and probably is going to start very simple like LinkedIn with just an ID verification. Then there will be an email, then it will upload a high school diploma or something like that. I think a lot more job boards are going to go that way, and that’s essentially also a new business opportunity because now think about the long-term perspective.
So, let’s say that all these ideas about automated recruiting actually happen. So, we have agents in the future which are able to [unintelligible 00:23:03] source profiles. Where are they going to get the profiles from? They’re not going to get it from LinkedIn because LinkedIn is closing. Xray is getting away. So, you cannot get them from the OpenWeb. You need to get them from somewhere else. And if we have platforms and job boards having their own, like closed, verified candidate databases, they offer an API on top of this that the recruiting agents can essentially go and source autonomously these verified candidates and deliver them straight to the recruiter. I think that’s where things will go into the next five years. But it also depends if technology is going to keep developing itself with the same pace as we see it today.
Matt: Yeah, absolutely. Very interesting times ahead. Alexander, thank you very much for talking to me.
Alexander: Thank you for having me.
Matt: My thanks to Alexander. You can follow this podcast on Apple Podcasts, on Spotify or via your podcasting app of choice. Please also subscribe to our YouTube channel by going to mattalder.tv. You can search all the past episodes at recruitingfuture.com. On that site, you can also subscribe to our monthly newsletter, Recruiting Future Feast, and get the inside track about everything that’s coming up on the show. Thanks very much for listening. I’ll be back next time and I hope you’ll join me.
[music]
