In the first episode of the Prospects Luminate podcast, Kate Morris, careers consultant at the University of York, talks about how employers are responding to the rapidly increasing use of AI by candidates completing job applications
The Prospects Luminate podcast is the place to come for in-depth conversations about the data, trends and research shaping how we think about early careers, employability, and the graduate labour market in the UK.
Subscribe and follow the podcast on Spotify, Apple Podcasts, Pocket Casts or wherever you get your podcasts. Alternatively, listen to more episodes here. You can also contact us with any comments, suggestions or feedback on the podcast.
In this episode, careers consultant Kate Morris discusses research by the University of York looking at what employers really think about candidates using AI in job applications, and how they are reacting to this fast-changing situation. A full transcript is available further down this page.
Listen now
Episode transcript
- Hosts: Dan Mason and Micha Smith, Prospects Luminate
- Guest: Kate Morris, careers consultant, University of York
00:00:05 Dan Mason: Hello and welcome to the first episode of the Prospects Luminate podcast. My name is Dan Mason and along with my colleague Micha Smith, we'll be bringing you a series of in-depth conversations about the data, trends and research shaping how we think about early careers and the graduate labour market in the UK.
Whether you're a careers professional supporting students in the transition from education to work, a recruitment expert aiming to hire the best new talent, or anyone else involved in the early careers space, this will, we hope, be an essential listen.
You can subscribe and follow the podcast on Spotify and in all the other usual places, and do get in touch by email using editor@luminate.prospects.ac.uk with any feedback or suggestions for what you'd like us to cover or guests you think we should talk to.
Now, in the first couple of episodes, we're going to be joined by some of the amazing careers and employability professionals whose research has recently been published on the Luminate website and funded by the Jisc careers research grant. You can find more details about that opportunity by heading to luminate.prospects.ac.uk and going to the About us page.
For this first episode, Micha and I talked to Kate Morris from the University of York about a really hot topic, the impact of AI, and in particular the research she and her colleague Claire McMahon-Harvey have done to find out what employers really think about the use of AI in job applications and how they are responding to the fast changing situation.
You can read the full report on the Luminate website. But before you do have a listen to this overview from Kate on some of the fascinating findings from their project.
And just one quick clarification of something I say while I'm introducing Kate, this research involved employers from around the UK, the only requirement being that they actively recruit in the York and North Yorkshire region.
We're joined now by Kate Morris, careers consultant at the University of York. Welcome to the podcast, Kate.
00:02:07 Kate Morris: Thank you. Hello.
00:02:09 Dan Mason: Hi, so your research is about AI and I know some listeners will want to devour everything they can about that subject and others have probably had absolutely enough of hearing about it, but I'd encourage everyone to stick around to hear from Kate, because this research is a really interesting insight into what employers really think about AI in job applications. And we're talking particularly about employers recruiting in York and the North Yorkshire region.
So, Kate, can I start by asking you to talk a bit about where the idea for this research came from and your motivations? Obviously, AI is a huge and current topic, but you came at this from a very specific angle.
00:02:51 Kate Morris: Yes, like most people working in careers, we're really interested in this topic. It's obviously something that is going to affect current students and graduates, but also moving forward, the impact on different industries. It's just such an evolving issue and really, the desire to research it came from a place of wanting to help students as effectively as possible.
So we had been actively looking into, you know, do employers mind if candidates use it, what's their stance? How does this work? And we were finding that both through kind of articles online, but also chatting to different employers. Everybody seemed to have a different opinion. There wasn't a consistent policy of don't use it or do use it. Embrace it. It was really mixed so it was making it really hard to know how what to sort of advise students.
So we really wanted to dig deeper into that to sort of find out what is actually going on, how are employers approaching this. Most of the research out there was looking at how candidates are using AI or how employers are using it in their recruitment but not about how employers feel about the candidate using it. So we thought oh, this is a really interesting gap in the research.
00:04:12 Dan Mason: Yeah. And I guess it's important too, you know, the fact that it's a gap is important because there's so many people putting so much stuff out about AI at the moment that, you know, you need to, you need to find that new angle, don't you?
00:04:23 Kate Morris: Yeah. I mean, I think AI is putting stuff out about AI. So it's kind of we wanted to kind of do some original research and really understand it better for ourselves.
00:04:35 Dan Mason: Yeah, absolutely. And so what made you look at the Jisc careers research grant as a way of funding this, how would you say the research aligns with the goals of the Jisc grant?
00:04:45 Kate Morris: We were really excited by the opportunity to undertake some proper quote unquote research, so obviously working with academics, they're always going on research leave and they're always getting grants for doing their research, and we've never really had that opportunity before. And I think careers, it's such an interesting area and such a difficult area to find up-to-date labour market information and facts about, as you say, there's lots of kind of stuff out there, but understanding what's true is difficult.
So seeing the Jisc research grant was like, wow, this is perfect. It's designed for people like us to support us in our work and also to sort of share good practice with colleagues. So it felt like it was going to contribute to the knowledge base of our industry. And it was a good opportunity to have some money because, who has money these days? So to actually have some money to do this properly rather than, we probably would have tried to do a mini version of it which would have, you know, been alright, but this was an opportunity to do it much more thoroughly.
We also knew that we wanted to recruit a student intern to support us with it, which was obviously going to give that person really good work experience, but it also allowed us to sort of share the workload a little bit. So yeah, we, we just thought it was a really great opportunity that aligned with the purpose of what we're doing and it made sense to just go for it.
00:06:16 Dan Mason: Fantastic. And we'll get right into that research in just a second. Just a quick reminder for everyone that once they've listened to Kate talk about the research, to go to the Prospects Luminate website, luminate.prospects.ac.uk to read the full report there. So Micha, should we get into the into the report in a bit more detail then?
00:06:36 Micha Smith: Yeah, sure. I was just wondering if you mind telling us a bit about your experience conducting the research, starting with like, what went well?
00:06:44 Kate Morris: Sure. So in terms of what went well, we found it really interesting to have a big project to undertake and it was fascinating to talk to the different employers. It was really great to have the chance to step back and think about what we wanted to find out, how we wanted to find it out and to feel like we were actively contributing.
We made the report. Yay, we finished it we managed it on time and it was published, and then we applied to deliver a poster presentation at an AGCAS conference, which was really exciting. We were successful at that. So we had our poster and we got to chat to lots of other AGCAS members about our work. I think we'll probably talk a bit more about the outcomes of the project later, but it's leading to ongoing other things let's say, I'll just tantalise you with that. It's leading to some ongoing interesting things happening as a result of it.
00:07:43 Micha Smith: OK, OK. Were there any challenges or anything you'd do differently next time?
00:07:49 Kate Morris: Yeah, there were challenges. So the flip side of never having done anything like this before was we didn't know what we were doing. So it was, we'd sort of not undertaken a proper research project before. So we didn't have research skills training that obviously a research student would have or academic members of staff. So for example, when we were making our survey questions for our questionnaire, we had to get all of that signed off by the university ethics committee, which was quite bureaucratic and I'm sure if you know what you're doing, it's much simpler, but we didn't, so that took time. But my colleague Claire led on that and that all worked out fine, but it was something that we didn't really understand how it worked from the starting point.
And also it was more tricky than we thought to actually get people to complete the questions. So it was that thing of like oh yeah, it's a really great idea but then getting people to give up their time to do it was a secondary consideration and we approached that in different ways. We made use of colleagues in our careers and placements team who work more directly with employers and had relationships, so companies who do a placement year, companies who do aa York internship with us, companies who are advertising on Handshake our platform. So they already had a bit of relationship and were able to send out information about it and try and get them to fill out the survey, that was fairly successful.
It wasn't enough, so we decided to do some in person. There were some in-person launches for the York internship scheme that we do where local employers who are interested in taking on a York student as an intern would attend these kind of networking events. And that was much more successful for us because we could actually go and talk to people and enthuse them about the project, explain more about it. We took our laptops with us and kind of pinned them down, not physically, but you know, we got them to do it while they were there and other people kind of agreed to do it afterwards and some people agreed then to do the more in-depth interviews. So we found that communicating directly with people about it was much more successful than a sort of colder email kind of thing. So yeah, that I would say, that is something that we learned how to get better results and get people interested in the research.
00:10:23 Micha Smith: Yeah, that makes sense. I was actually wondering about that because I know right now it's particularly challenging to get people to answer surveys because a lot of people are just fatigued with surveys, especially since COVID. I was also wondering, did you encounter any challenges in getting employers to engage with the topic of AI and recruitment particularly? Or was, were people who were willing to respond just open about talking about that?
00:10:50 Kate Morris: They were very open in talking about it and very, very interested in this topic as a whole, but also for, you know, quite a few of them they hadn't really thought that much about the candidate side of it. They were thinking more about how to use it in the workplace, what it might mean for their business. They hadn't necessarily thought about, you know, how people are using it for job applications. I'll caveat that with the large employers had obviously thought about this topic and are being impacted by it more, more overtly but yeah, no, there was no problem in getting people interested in the actual topic itself. They were ready, willing and able once they got it, and they understood why we were doing it.
00:11:30 Micha Smith: OK and I see that you used a mix of qualitative and quantitative methods. What would you say that combination allows you to uncover that maybe a single approach might have missed?
00:11:42 Kate Morris: It was so useful in allowing us to analyse trends and also dig down into the detail, so a lot of research online is showing that you know, there's been this huge surge in applications because it's so easy to pump out an application now using generative AI. So it looks like everyone's using it and it looks like employers are complaining about it and having problems with it. And actually, because we were, we had both large employers but also lots of SMEs completing the survey it allowed us to understand, actually this is different in different contexts. So to have some overall trends, but to have some specifics as well. So for example if you just looked at the overall trend, it looked like lots of employers did not have a policy about candidate use of AI, but if you drill down into sort of the detail, the larger employers did have or were working towards developing a policy. So it was kind of it allowed you to have that more, more nuanced approach, and we also found that having the longer in-depth interviews gave us a great opportunity to get some great quotes which bring the research to life. So you're not just 'here's another chart', 'here's some numbers'. Lots of us learn different ways and take in information in different ways. Some people are going to really like the detailed data, but other people, it's more the storytelling that's going to resonate with them. So I think having that combination of both the qualitative and quantitative research is kind of helpful when it comes to writing the report and sharing the findings with different audiences.
00:13:31 Micha Smith: Yeah, and that extra information kind of brings it brings the data to life, doesn't it? Like, it allows you to not just see a stat but also be able to contextualise it, like what might this stat actually be saying. I also see that you used, you interpreted the data using thematic analysis as well. Were there any themes that emerged unexpectedly or that challenged your initial assumptions through the thematic analysis.
00:14:02 Kate Morris: Yeah, I think certainly the policy side of it, we were quite surprised about the lack of policy and the lack of guidance. So a lot of employers were quite negative or had negative perceptions about candidates using AI. But it was interesting that they weren't telling candidates how they felt about that. So that was that was quite shocking and interestingly, looking also at their awareness of candidate use of AI. So again, as I said the larger employers were pretty clear that it was coming through on applications, but a lot of the SMEs hadn't noticed it or thought about it. So it we kind of thought either that means they are the candidate is doing a really good job or the employer is just not really understanding what it might look like, an AI application might look like.
We found that as well there was a real lack of understanding of the impact on diversity. So you're in the sort of university sector, there's a lot of emphasis on making information accessible, teaching accessible, documents accessible, better understanding of neurodiversity and other kind of issues like socioeconomic background and how that might impact on academic study and career development. And there are potentially lots of positives that using AI has on those things, but employers had really low awareness of that. So again, looking at it through the lens of some different things that we're interested about in the context of our work and applying that to the research as well and including questions to kind of surface those issues was really useful. And brought out some different things.
I think for me, I was, the thing that I was most shocked about was that neither party really understood each other's viewpoint, so the employers weren't really clear on how the candidates felt about the use of AI and applications, but the students and candidates, if you like, were really not clear on the employer perspective on them using AI in their applications and also are really not clear on how employers are using AI in their recruitment as well. So it really felt like there was, there's a real gap between those two parties and this research and university career services I personally think are in a really good position to bring those two perspectives together and kind of facilitate dialogue between them.
00:16:43 Micha Smith: Yeah, for sure. Moving on to some of the key findings. Firstly, before I ask any questions, are there any particular findings that you'd like to highlight?
00:16:52 Kate Morris: I think that one about the lack of understanding of both of both perspectives was a key finding and I think that gives us a really interesting jumping off point for future research. I think a really core key finding was that lack of awareness of how it impacts on diversity. So either through you know socio-economic background, where if you have more money you could maybe pay to use better versions of generative AI, which is going to allow you to do better applications. Similarly, if you have and you have a neurodiversity, which makes it more difficult to structure your thoughts or articulate yourself in writing, obviously, as a tool, generative AI can be really helpful, and it was surprising that employers hadn't, where they were thinking of just like a blanket ban, they hadn't thought about how that might disadvantage some groups.
I, there's something that I found really funny as a key finding was whilst a lot of employers either didn't want people to use it or use it in a really limited way, they were actively using themself in the workplace and they want, they want candidates as new employees to be able to use it to good effect in the workplace and take advantage of these new tools. So it was really interesting, kind of paradox between you want people who have the skills and the confidence to engage with AI. But you don't want them to utilise that, those skills and interest as part of their application. So it was kind of like that's just such a weird thing. Again, maybe the dialogue and an understanding of both perspectives would really help to address that.
00:18:39 Micha Smith: Right. So it's not like a general fear of AI, it's just maybe more of a fear of not, not being able to screen candidates in the most effective way and knowing their skill levels if they misrepresented their skills.
00:18:54 Kate Morris: Yeah, absolutely. That that was the, a kind of a key thing. They were either concerned about misrepresentation. So they were going to end up with people who couldn't do what they said they could do, and also a lack of authenticity. So not really understanding if this candidate really is interested in them and really understands what they do. So yeah, I think it was, they were, how do we read this? How do we assess candidates if everything's the same and everything's kind of a bit fake?
00:19:25 Micha Smith: Yes, surely. And just touching back on the idea that many weren't aware of how AI can assist applicants who may have some type of neurodivergency. What are the risks of developing AI recruitment, AI related recruitment policies without considering equity and inclusion?
00:19:48 Kate Morris: Well, interestingly, they were sort of seeing AI as like them using it, the employer using AI to help sift applicants as a way of avoiding bias and helping reduce discrimination. But obviously if you, who's programming those things and how they're operated, it could increase bias. And also if they are kind of determining that if you've used AI, then you're not authentic and you're not a good candidate and they're going to sift those people out, whereas actually those people might have been using it as a tool to address, you know, to offset a potential disadvantage that they have because of disability or background or whatever that is going to lead to increased, potentially lead to increased discrimination and more problems for, you know more struggles for candidates in the workplace and also goes against probably most of their own policies in terms of increasing diversity and equality and those kind of things.
00:20:55 Micha Smith: OK. Can you go more into how career services can help bridge the gap?
00:21:03 Kate Morris: So there's a few ways it, one of the questions that we, or two of the questions that we asked the survey respondents was whether they felt that universities should provide training on applicants to use AI effectively in applications and also whether university should provide training on AI and how to perhaps, you know, upskilling that for the workplace and overwhelmingly they were positive towards that. So it was like over 85% in both of those questions said that, yes, they think universities should be providing training for that. Now obviously, that doesn't necessarily mean it always has to be the careers team doing that. There's multitudes of different potential places. It could be academic skills. It could be as part of your module. It could be a combo of all of those things. So there's something around actively training people on it.
But I'm also interested in this conversation side of it. So we're participating in, you know, obviously undertaking this research was a first step, going to the conference. We're going to be doing a mini conference at York in a couple of weeks time with academic staff to raise awareness of this, we really are kind of interested in organising maybe a round table discussion, so getting some employer representatives and some student representatives and bringing those two parties together, we have an AI working group within the careers team and there's a wider university AI working group as well that my colleague Claire is a part of. So I think it's a multi, a multi-pronged approach really. I think it's awareness raising, but there's also some really interesting practical things that can be done around training and awareness, both on the employer side and on the student side. Oh, I'm running a training session for students this semester as well on AI and your future career, so I'm going to be looking at you know, what's the employer perspective? How does this work? How is it impacting on different industries? But I'm also going to demo how to use ChatGPT to make an application and refine it, how to assess it. How to bring it into your authentic voice, so you're using it as a tool rather than just letting it churn out a stock answer. So I think there's lots of different ways that you can engage with it and kind of help that conversation.
00:23:39 Dan Mason: If I can just, sorry Micha, just come in there and go back to the point you're making about. Or the finding that, the vast majority of employers favoured universities providing this AI training and you can definitely make the case that universities are a good place for that to happen. But do you think there's also an element there of sort of employers passing the buck a bit and sort of maybe saying ohh yeah, could you fix this for us because we're struggling to know what to do and sort of saying to universities, yeah, could you solve this problem?
00:24:13 Kate Morris: Yeah, I think so. I think they kind of, nobody knows what we're doing right. We're all just learning about this. And I think that they want the students to come ready to engage with this and they want they are expecting probably the students to be better at it than them because they're, you know, born in a tech world, etcetera. Yeah, I think you're right. And I think ideally it would be collaborative so, again, the idea behind sort of a roundtable and developing training would be to sort of I think it would be very useful to have the employer voice within the training that you did to help really ground the training in what is needed in the workplace and how is that evolving. So something that we already have been doing for quite a while is when we do panel sessions about different topics or different professions and we get, you know, graduates or other professionals coming in to talk about their work as part one of the panel questions where you would usually ask is you know, what's the impact of AI and your industry. So again trying to get them to give an insight, not necessarily of the whole sector, but certainly on their role, their business, how is it already impacting. So just to keep asking all the people we meet about this and sharing that with the students, that in itself, it's the employer doing a bit of work there because they're giving the intel and the perspective, but we're kind of joining that up with sharing it with students and helping the student engage with the topic and understand it a bit more.
I think in terms of more formal training. Yes, you could say it's down to the employer, but obviously a lot of students we work with don't necessarily know what industry they want to go into or what employer they want to target. So I think it in the in the employees defence, they may not get great attendance if they just did a standalone session on use of AI in that particular sector.
00:26:13 Dan Mason: Sure, sure.
00:26:17 Micha Smith: Just one more thing before I ask about the recommendations. So you did find that larger companies were more likely to have an a policy on how they would like students to use AI. Were they also more likely to be actively engaging with students on their perspectives on AI? Or was that like not present at all?
00:26:40 Kate Morris: Sorry. So just to clarify, clarify the larger employers who did have a policy where they sharing it more effectively, is that what you're asking?
00:26:48 Micha Smith: Were they engaging with student perspectives on AI, or were they also like unaware of how students felt?
00:26:56 Kate Morris: Yeah, that's a really interesting question. So a lot of them were really unaware still of how students felt and it's, I can see it from their perspective. So I mean, they were aware that people were using it, so they weren't, they didn't have the same lack of awareness and 'oh we didn't, we haven't noticed it coming through' kind of thing because obviously they're getting a big increase in applications and some of them are looking very similar slash identical. But I think employers have maybe forgotten how time consuming and soul destroying it is to make an application so students are working hard, they're studying, they're doing their part time job. Applying for a job if you know if you're going to do it well, takes hours and hours and hours and you never hear anything back and you know, I see students in appointments who are kind of really fed up. They've been trying their best. They've been putting effort into these things. They don't know where they're going wrong, obviously any university careers service is going to provide resources and training and how to do those things, and so students can access that, but you can still kind of do your best, do all the right things and still not get anywhere, and you know that might be because you don't have as much experience as the next person or you know, whatever. Or just there's millions of people applying and you just slip through the net. But I think it's totally understandable as to why you would try to make that process more efficient and take advantage of things that are going to speed it up. Help you potentially be more successful at it. So I think it would be helpful for employers to think about you're making this a really onerous, clunky, difficult, time consuming, stressful process and you give nothing back. So it's pretty understandable that people are going to try and make that easier for themself. So there's potential I think, this is my, m view long term again, it would be nice to see. Look how can you either give better support and feedback to candidates, or how can you simplify the process to make it more authentic and less stressful on both parts? You know, for the candidate and for the employer.
00:29:10 Micha Smith: So your report outlines multiple recommendations for universities and career services. Which do you think is the most urgent or impactful to implement, right?
00:29:22 Kate Morris: I think the training side of it, I think really helping students understand what is potentially going to disadvantage them by just simply using AI and not refining the the outcome to that. I've literally met students who have, 'I've applied for 250 jobs and I haven't got anywhere'. And then when you actually have a look at an example of what they're doing, it's no wonder, but they don't know that it's, that's bad because they're like, oh, the computer told me to do this. This is perfect. This is what it's advising me to say. So I think helping students understand actually this can be a really useful tool, but this is how you would go about doing it. And also it's not going to be acceptable by all employers to really make sure they check if it's allowed, what the policy is, that kind of thing. So I think really helping the students understand if they are going to use it, how to use it to good effect. But I really think the other side of that is this piece around working with the employers to sort of say, look, please let your candidates know what your policy is. And be fair, you know, it could be a really useful tool and it you don't have to necessarily have a blanket bann against it. So I think, yeah, it's this piece around how to use it if you are going to use it well. But also being clear with candidates, what's your view on this? Can they use it? Can they not use it?
00:30:51 Micha Smith: And are there any plans to do, actually you kind of spoke on this already, you said a lot of interesting things really about follow up research that you plan to do. Do you think that are there any areas that you think will benefit from further exploration by others as well?
00:31:10 Kate Morris: I think it would be really interesting to look in more detail at how employers are using AI in recruitment so, you know, we've heard a lot of scary stories about, you know, AI assessing their candidates and just kind of eliminating lots of people. But there was an interesting article. I think it was, yeah, it was on the Luminate website actually looking at, oh, a lot of it is automation rather than AI as such. So there are going to be some standard ways to eliminate certain candidates, say on, they haven't achieved the right grade or, you know, whatever rather than the more nuanced information around how someone's articulating themselves or how they're expressing their motivation for the job, that kind of stuff. So I think there would be, it would be really interesting to look at employer use of AI in recruitment, in assessing candidates.
I personally am just so fascinated on how AI is impacting on actual people's jobs and different sectors. So, and I think that could just be like an project, you could just continue doing for a long time. But that labour market information side of it. This is so fascinating. We're at a really unique kind of potential turning point in history. It's like, how is this impacting? What will our jobs look like in five, ten years time? Will we still have jobs? Let's hope we do. And I think that's a really, that's a really fascinating piece because it's constantly evolving. And just again taking that opportunity to talk to different people from different sectors and keep a live up-to-date insight into what's happening is important, but I think it would be really cool to do some more in-depth research about that, whether it was on a particular sector and the impact or whether it was broader across a range of sectors. Yeah, I think there's, there's so many interesting ways that this could be sort of developed further.
00:33:12 Dan Mason: Brilliant. Well, it's been an amazing overview of the report and the research. Is there's anything else you wanted to add, Kate?
00:33:20 Kate Morris: I guess, just thanks to Jisc and Prospects for the opportunity because we really did enjoy that. It was, you know, it was time consuming, it was busy. It was a big long project, but we really enjoyed having that opportunity to take a step back and, you know, research the topic in more depth. We really enjoyed the opportunity to recruit a student and we had quite a lot of students apply for the, for the research assistant job that we created for the internship and it was amazing to provide a student with real work experience for that. We really enjoyed the challenge, the personal and professional challenge of undertaking research, and it's been really exciting. So we just really encourage colleagues out there to go for it. If you think you've got an idea, don't be afraid. And I would also say it was really useful for Claire and I to work together on it. She was the project lead, but we could really support each other. So she, when she was really busy, I could take stuff over and vice versa. So I think maybe you know you could do it alone if you, if you have the time, but it's really nice to do it in partnership with another colleague and that the team the teamwork makes the dream work kind of thing. It really did help. So yeah, I would say go for it. But think about realistically, you know, can you do it all on your own? Is there a way to have an opportunity to work closely with a colleague on this?
00:34:45 Dan Mason: Brilliant. Well, thank you very much for doing the research and for putting out there through Luminate. And it's been brilliant to talk to you today.
00:34:54 Kate Morris: Thank you for having me. It's been really lovely experience.
00:35:23 Dan Mason: That's all for this episode, I hope you found that discussion as interesting and valuable as I did. Micha and I will be back to chat with another of the researchers funded by the Jisc careers research grant in the next episode, so I hope you'll join us for that. In the meantime, remember to subscribe and follow wherever you get your podcasts, head to luminate.prospects.ac.uk for all the latest updates there and send us your emails to editor@luminate.prospects.ac.uk. Thanks once again to Kate and to you for listening, and I'll speak to you soon.
End of episode
Note on transcript
This transcript was created using a combination of automated software and human transcribers. The audio version is definitive and should be checked before quoting.
Written by
Was this page useful?
Thank you for your feedback