AI-drafted emails from prospective PhD students

[GCAcademic]

I'm wondering how other academics on here are handling enquiries from potential PhD students who are clearly using AI to draft their emails? Some of these are clearly phishing emails, while others seem to be genuine but the email is very obviously AI-drafted (they all adhere to exactly the same formula).

I've got to the point where I just want to ignore all of them, but I worry about missing a potentially strong application where the prospective researcher has simply used AI because they are unsure or nervous about how to approach a potential supervisor, particularly where there are cultural differences.

And, yes, I've only done a simple tweak to the AI-suggested title for this thread 😂

At my husband's work they don't consider these. At all. They bin them. More than enough to choose from.

The problem is that they nearly all seem to be like this now.

Surely this makes your job easier?

You can weed out those who can't even be bothered to write their own application. Shows how much effort they would put into anything else doesn't it?

The ones who try harder should be rewarded imo. If that's a smaller pool, then so be it.

They find their people anyway sorting.

“AI-written emails are deleted immediately without consideration. You are welcome to contact me again if you wish.”

I like this, thank you.

It’s hard. As academic should, by definition, be at the forefront of researching and implementing innovation. These students are.

I wonder if there were analogous “conundrums” back in the day when there was a transition from hand written to typed application letters…. And look where we are now.

Oh dear yes.

Sometimes they suggests topic nowhere near my expertise. Those I forward to my DPGR. If they’re in my bailiwick, I respond and have a standard email with details about fees etc plus ask for a 1500 word proposal.

If I hear back then good. If I don’t, also good.

My DPGR will often tell me not to bother to respond as the applicant has spammed the whole department.

Hmmmm, I don’t think innovation is the point here. The AI emails I get are obvious - they haven’t done their research as they’re approaching me to supervise on topics outside my specialist area. Not what a PhD enquiry should do. It suggests laziness or desperation not being at the forefront of one’s field. They can’t even be bothered to search me on Google Scholar!

Is this issue though that you can't prove that someone has used AI and if they have, to what extent. Unless your uni has a policy of immediate rejection on suspicion and a clear public policy stating no AI at all, are you not potentially putting the uni at risk of a discrimination claim if you follow some of these suggestions - it's very possible that someone with a disability for example may use technology to assist them.

Yes, partly this. There is no university policy. The last two of these emails I've received are from individuals with good MA degrees who could be strong candidates but the fact that they've used AI to draft a generic email makes me want to ignore them. I would need to undertake a fair amount of due diligence to ensure that they have written the proposal themselves and, frankly, I don't have the time or inclination for that.

On the other hand these two particular cases are from other countries and I've seen enough on here and on social media to know that candidates are nervous about approaching prospective supervisors, particularly when they're not familiar with the academic culture in the UK, so I could kind of see that they might use AI to help them make that approach "correctly".

When I applied for my PhD position, a key aspect of it was how I critically engaged with the project (there was already a framework for the one I’m doing). Critical thinking skills and original thought were central to my proposal. I am far from perfect as a researcher I’m sure, however these aspects surely should be integral from the outset.

I was wondering about this. Do they know they're not supposed to use AI? My partner has recently applied for a position where it was a requirement that he use AI to help with the application. (For an AI-related job, though).

I don’t think that’s particularly analogous at all really.

But if it was, then an email that is obviously AI is akin to a badly typed letter full of typos ie they’ve used the new technology poorly. It’s perfectly possible to use AI to help in writing a letter in a way that improves it but is not obvious.

Of course there is the issue that you might falsely accuse someone of using AI.

Just to be clear, it’s not the proposal I’m talking about, it’s the initial email. I haven’t received an actual proposal yet, just the enquiry asking if I’m taking on supervisions and could they send me their proposal on X topic.

Just wait until you get the proposals created by AI…! I’ve had a few of these. They look really plausible at first blush but of course if you know anything about the subject it becomes clear AI has hallucinated their references for them. One even made an argument based on a collection of documents which was under embargo at the time (you would only know this if you approached the specialist library where they were being held).

I actually did write back and explain exactly why I wasn’t going to take their proposals further and they had the good grace to confess to using AI. It didn’t change my mind though.

Edited for a typo.

It’s a race to the bottom. I think it once people think everyone else is using AI there is massive unseen pressure to be shiny and corporate and so they use AI as well. We are all becoming brainwashed to think that AI is going to produce something better than we can produce.

I think in your shoes I’d have a standard form response saying they are welcome to submit a proposal and setting out what it should include. Also making the point that use of AI here is unhelpful- you want to hear their own ideas in their own words.

I would try to resist seeing it as an attempt to cheat. Students these days are encouraged to use Ai so they may think this is what they are supposed to do

^This and use AI to write the response.

@664eek may I ask why students are encouraged to use AI, and by whom? I'm not in the university field so on the outside looking in.

I also can't believe it's now thought impossible to be polished and professional without it!

Those AI posters which are so full of too much information and all look the same leave me cold and I actively avoid them. I know they save time for the maker but they aren't working (for me at least).

Have our standards really fallen so low?

I almost must the very blunt/rude Dear Smith emails, although not the ones that were asking for phds on topics 3 subjects over. I work in a non-English speaking country, so most students use AI for translation (as do I) which i think is fair, but there still needs to be some personalisation and human thought structure. I think ERC grants now have to be declared AI free.

Using AI to draft an email where it's really obvious that you've used AI is not being at the forefront of innovation.

I would argue such a basic use of AI in the face of lots of freely available advice on how to use AI while maintaining your voice is actually behind the curve now.

The issue is if you ignore an application from a person who used AI as a reasonable adjustment, you’ll be in a load of hot water

not for PhD you won't.

I just ignore these, OP. I get at least 2 a day now and don't even reply. straight to the delete button.