"AITA for firing someone because they used AI?"
I feel pretty conflicted about this and anyone I’ve gone to for advice is pretty 50/50 on the matter. So, I’m turning to potentially brutal internet strangers because the last time I came for advice, it was surprisingly helpful.
For context, I run a small marketing agency of 8 people, including me. When AI like ChatGPT started to get really widely used, I made a decision with my staff that we would commit to not using AI in client work. So everything written, designed, or photographed by our agency would be 100% done by us. We communicated this with clients and essentially, made it a brand promise.
So about 3 months ago I hired a recent college grad for an entry level social media position. In the job listing and the interview, I made our commitment to no/low AI usage very clear. It was framed as if you want to use it to optimize your calendar or your own personal to-do list, I don’t care, but any content created by or associated with the company, will be made by us.
He agreed to this and it was actually written in his contract when he accepted the job. About a month in, I was reviewing some of his work on a project and I was a little suspicious. There were very commonly used sentence structures, M dashes, and pretty boilerplate language for the industry that he was writing for.
I did check his search history, and as an admin, I could see that ChatGPT was in the history. I didn’t dig any deeper, but did slack him and ask him to stop by my office when he had a second.
When he came in, I said that I just had a couple of questions about the scheduled posts for said client, which I did, and then right before ending the conversation, I just flat-out asked him if he used AI for any of this. It was quiet for a second, but he did admit that he put it in ChatGPT to check spelling and grammar.
I told him that the spellcheck on Google Docs is pretty good at catching things, and as far as grammar goes, he could ask anybody in the office, including me, to proof things. I actually encouraged it because that’s a muscle that I myself want to keep using, and it’s great for everybody else to have that mindset as well.
Then I reminded him that in his signed agreement, he acknowledged that we do not use AI in any writing or creative work, I asked him to not let it happen again and we ended the conversation.
I did admittedly keep a little bit of a closer eye on his work and checked his search history a couple of times in the next two weeks but dropped the issue after everything was looking good. Not too long after that, our copywriter brought him up in my 1:1 with them. We’re at the end of our normal meeting and just chatting and she asked if I’ve noticed anything about his work.
I honestly didn’t think about the AI thing right away just because I thought it was a non-issue now. I asked her what about it and that’s what she told me that she saw ChatGPT up on one of his tabs when he was showing her something on his computer and so for about a week she’s been paying attention and noticing some small nuances that have made her think he’s using it regularly.
Considering she’s a copywriter and I’m not, I trusted her instincts on that even more than mine. So I gave her a quick overview of what he and I had already talked about, and she threw out that he could either be in an incognito browser or be doing it on a personal device, basically, I could be being naïve and giving him a little too much credit, and the problem might not actually be resolved.
His 60 day review was the same week, so in it, I brought up the topic. He kind of clammed up and apologized for using it in the first place, but said that he has not used it since, and I just didn’t fully believe him. I didn’t grill him, but I think I did go a little bit harder on him in this review then I did and his 30-day.
But in my defense, I do that to everybody because you should be held to a higher standard two months into the job than you were one month in. But given everything, I did give him a kind of false assignment. I took an old brief for an existing client, it was a project we ended up tabling for another time.
I gave him the brief with specific deliverables and asked for a draft by the end of the week, which is about how long it should’ve taken with his workload, experience, and what I was asking for. I got it back in two days. Which, yes, is possible, and usually a boss would be thrilled about.
But then I ran it through an AI checker. Which I know is using AI to check for AI. I didn’t like doing it, and it’s pot calling the kettle, but I needed some level of proof. It came back as pretty clearly AI. So I put in his work from the past two weeks, also AI.
I documented all of this, brought him into my office, and told him that effective at the end of the week, he was let go. I did give him the names of multiple contacts in the area who I was connected with and told him I’d gladly be a reference without mentioning any of this. But ultimately, he repeatedly broke an agreement and I wouldn’t tolerate it.
It’s been a week and I feel horrible about it, but I also don’t. Not to sound like an old fart but younger generations are losing skills and originality in the name of false convenience and their own impatience. I don’t want that in my company that I’ve built and it’s not fair to my other employees who, as far as I know, are holding up their commitment.
But my conscious and some people I’ve confided in about the decision ask if I was just too hard on him or if I didn’t have enough evidence or if I’m just stuck in old ways. So, where are you at with this?
Here's what people had to say to OP:
venom029 said:
NTA since you were clear from day one, put it in writing, and gave him a warning after catching him the first time, and he still kept doing it. at that point it's not about ai vs. no ai, it's about repeatedly ignoring a direct policy he agreed to. you gave him multiple chances, and he chose to keep breaking the rule.
wamimsauthor said:
As a copywriter of over 18 years NTA. However, I will say I’ve been accused of using AI when I did not. It depends on the type of program you’re using to check. I don’t use AI because why would I when I’ve been writing almost two decades without it?
NYCStoryteller said:
NTA. It's part of your contract and brand promise, he violated it repeatedly AND he lied about it.
Insulo said:
NTA. But don't trust "Ai detectors". There's no good AI detection in the marked yet. They all make too many mistakes to be reliable including saying text from before 2010 is AI what is impossible. Check another way for AI. But AI is terrible at detecting AI. A lot of false positives because it thinks giving a "Positive result" is better. It's not actually analyzing anything, it's just answering what you want to hear.
Loud_Bodybuilder546 said:
Hell no NTA. It was his wake up call. It’s entry level, he will find something else. I’m so over grads in recent years not being to do anything or think for themselves or be creative because they just use AI.
TinMachine said:
NTA - however I would be very wary about AI checkers - I don't think they would be reliable, especially for copywriting. But I do think you can still generally get AI vibes and experienced professionals can trust their own instincts on it.