Fed up of lazy AI use in recruitment

[DerelictWreck]

AIBU? Are others finding this?

We use an online recruitment system at work which is designed to reduce bias - we can only see how candidates have answered work based questions which are designed to assess their ability to do the job. We don't see any personal information or CVs.

As part of the process the candidates are told that use of AI will not help them and when they submit, they have to sign a declaration that it is their own work, not generative AI. Any AI work found will mean they are automatically dismissed from the process.

But over the last year this has so rapidly scaled that I'm now looking at over 50% of applications copying and pasting from AI chatbots. It's blatantly obvious, gives poor answers, massively slows down my recruitment processes, and is a complete waste of the candidates time as they get auto-rejected! I'm at the point where I'm also going to start blacklisting their names from future recruitment, and sharing said list with our sister companies (part of a large network).

I don't understand why they do it - using it to get ideas and editing it I get. But copying and pasting - how stupid do they think we are?!

But @TheBeehive the example I gave was where people have to sign that they haven't copied from AI. So they can't think it's not cheating.

As I've said I've no issue with people using it intelligently as part of their prep, it's the lazy copy and paste "I'm too good to try" / "the recruiters are too stupid to notice" I don't like.

It’s also known that they are more likely to flag non native speakers as AI. Which would potentially be racial discrimination.

thats a fair point, could the issue be the devil in the details eg ive not used ai for all of the answers vs i only used ai for part of the answer so techinally ive not cheated ?

apart from those that just copy and paste,

@DerelictWreck Note, I didn’t say all questions are problems and specifically called out that I’m assuming these are written. I’m also assuming that they are scenario based.

Personally, unless the job is something like journalism, editorial etc rather I think skills based assessments are probably more fair, even though I know most people hate them. This is just a layman opinion, but why go through the trouble of blind recruitment only to reintroduce bias? I recognise that there will never be a completely biased free recruitment process though.

I don't understand PP saying the applicants using AI is "a waste of my time" and "blatently obvious"

That means it's incredibly easy to exclude those applicants. So how is that a waste of time, 1/3 easily excluded. Sounds like a win to me.

It’s Civil Service. I’ve given up applying to them, I clearly don’t have the type of brain they’re looking for, despite only applying for jobs I’m well-qualified to actually do.

@DerelictWreck but on topic YANBU. If somebody is happy to fill out an application, they should adhere to the requirements.

I realise these things take time, and sometimes the data isn’t in place to track this over a number of years, but have you found that it’s led to a measurable increase in diversity?

@DanaBarrett I have literally been thinking this must be public sector, but I think somebody above said they do the same thing in their company.

We don't use such tools so I'm not questioning the veracity of this, but it surprises me in a way because I usually find the written English in AI posts to be impeccable. The content not so much....

We have applications where we suspect the use of AI but we can't prove it so we just score them as usual. Generally, the AI content is pretty shit so it doesn't get shortlisted anyway. If someone actually manages to generate consistently high quality answers from AI, then fair play to them... they're probably worth an interview!

We are not public sector.

Well, AI is an available tool so this was bound to happen. In your shoes I'd revamp your application process because it clearly isn't working. Its some years since I applied for a job but I still remember rolling my eyes at applications which required a cover letter + CV + application form or the dreaded long written statement as to why you want the job and what skills you'd bring to the role read the bloody CV and cover letter

AI is here to stay it'll be like chasing your own tail thinking you'll get on top of it. & as pp's have said, those tools claiming to be able to spot AI aren't reliable. Best off going back to CV and interviewing applicants you like the sound of

The Civil service recruitment process is so limited, I worked for the most central office in central government you can think of in the late 00s, I recently applied for a data job that I am very capable of doing having 6 years in experience in such a role and was rejected, when I read the feedback it was astonishing and I am utterly perplexed as to what they are expecting for £42000 a year. My DH who is on closer to 90 in the private sector was in the background and assumed from their questions about overseeing three departments they were actually in need of a SCS!

Bias and AI is quite well documented, because AI is dependent on the data, it’s given so can reflect the biases of the person/people involved in it’s creation. This was a big thing a few years ago when Google fired one of their senior AI staff because she co authored a paper about it.

This is the thing it is deemed fair and best practice but actually is inherently bias. What about a conversation for recruitment, ascertaining candidates actual ability to do the job.

Well, that's essentially what an interview is, isn't it? And realistically, it isn't possible to interview every candidate for a role... there has to be some kind of sifting process.

I don't think it will ever be possible to eliminate all unconscious bias from the recruitment process, but that doesn't mean that we shouldn't try to make the process as fair as possible.

It's not fair with certain software though.

You say it as easy as that but many recruiters don't do that based on a meritocratic system as they have eliminated people via their biased software.

Exactly.

Yes, I suppose some recruiters do eliminate people using software, but that's not really what this thread is about, is it?

We don't use software to sift or screen applicants, and it doesn't sound like that's what the OP's employer does either. Except that they seem to be using some sort of tool to check which responses have been generated by AI? As I said above, we don't bother with that as I think the AI answers are usually pretty crap anyway.

I wouldn't personally trust a computer to do the screening, but I can see the attraction of doing this if organisations are routinely getting hundreds of applications for every job... which is pretty common if you allow people to apply by just sending a CV.

Bravo @Pupsandturtles, could not agree more.

As someone who spends a lot of time rewriting things that haven't actually answered the question that was being asked, I can see AI as a first step for processes like yours being used to weed out the ones where the answers are completely irrelevant. [Grin]

I think the questions asked are pretty important but I honestly am not surprised that you are finding better, and more diverse, candidates. I think it's a huge improvement on "ooh, look, she went to s shitty University, no way can we hire her" of the old days.

Out of interest, could you provide an example of the spec of question and sector. I haven't applied for a job in 10 years so my knowledge of these is all theory and am interested in how it works in real life.

Same! Also, fed up of companies using ai to sift applications and complaining when candidates use it as well.

Purely from a common sense/time/cost perspective I can’t see why adding another stage into the process(running the ‘is it AI’ check) is worthwhile. Aren’t you risking screening out candidates that MIGHT have been suitable? They just didn’t have the time to answer loads of questions. You’re in competition with other employers after all, however reputable you are.

As a recent, reluctant job hunter I’ll say what I wanted from the process: a clear ad, salary/benefits visible, a quick application process with transparency about the interview process(how many interviews, presentations, when can I expect feedback).

If you want examples of written work, or case studies, get them at the interview stage. You’ll get better results when you’ve engaged people’s interest and they know they’re shortlisted surely.

The thing is, if employers are attracting plenty of high quality candidates for the roles that they need to fill, they're not going to be that bothered about what you want as an applicant. I am not saying that to be disrespectful - we have all been the applicant at one stage or another - but merely stating it as a fact.

Yes, we are technically "in competition" with other employers for the best people, but if the evidence suggests that we are already competitive enough to attract the kind of candidates that we need, then there is no need for us to bend over backwards to meet the expectations of candidates if it doesn't suit us as an organisation to do so.

And yes, there is always the possibility that we might miss someone who would be good at the job (at least technically) but can't be arsed to answer loads of questions. Honestly, I'm fine with that. Candidates who can't be arsed aren't really what we're looking for.

I completely agree with you regarding having as much clarity and transparency as possible around the role, the selection process, the timelines, salary and benefits, hybrid/flexible working options etc. This is in the interests of both recruiters and applicants. I don't want to waste other people's time any more than I want them to waste mine, so being open and upfront about all of this stuff is very sensible.

However, as an organisation, we will stick with the shortlisting processes that are demonstrably working for us in terms of overall quality as well as increased diversity. If our process puts people off because they're looking for something quicker and easier, then so be it.

Don't want to share too much potentially outing info, and the questions vary according to the type of role. Typically, we're looking to assess skills, knowledge and insights as well as attitudes and values. We generally use at least some scenario questions but we have others which don't follow that format. What we have learnt is that the process only works effectively when we get the sift questions right...it's a bit of a skill to write them and we have all had to learn how to use the system in an optimal way.

I do find it amusing that the industry that ‘pioneered’ ATS which is a primitive AI to ‘screen’ applications that have taken people hours to write as they couldn’t be arsed to read them are now whinging when candidates dare to try to use the same technology to get ahead. Apply for jobs takes hours and applications have no idea what ATS ‘key words’ you’ve programmed so yeh some people turn to other things they hope might get them passed your AI!!!

@MrsBennetsPoorNerves I think ‘can’t be arsed’ is a bit unfair. It isn’t always a case of ‘can’t be arsed.’ Sometimes it’s a case of ‘there are only so many hours in a day.’ Particularly if you’re prioritising diversity- what about women who have kids and do the lion’s share at home, what about people from lower socioeconomic backgrounds who might already be working multiple jobs?

it’s great that you have such a high opinion of the organisation you work in. In my experience of working in the third sector though, I’d say every HR department at every firm I’ve worked for would describe themselves in the same terms. To be blunt- from a candidate’s perspective, your org may not be as special as you find it to be, and the competition may be fiercer than you expect.