my dishonest employee blames technology when he messes up

by Alison Green on November 6, 2024

A reader writes:

How do I deal with an employee, Miles, who is dishonest but gets emotional and doubles down when confronted with technological proof?

For instance, he’ll tell me he submitted a work product when the timestamps on shared documentation (which are visible to both of us) show that the work wasn’t completed by the deadline.

Separately, I’ve suspected that he’s using generative AI to complete work, and multiple AI detectors — which I acknowledge are imperfect — also say large parts of his work are AI generated. (It may be fine to use AI in other circumstances, but it violates our policy prohibiting using AI to write external-facing documents, which he’s responsible for.)

If I confront him, Miles gets defensive, frustrated, and visibly emotional, denying that he missed the deadline or that he uses AI or whatever, and insists he doesn’t know why the timestamps/AI detectors would say that he did. I feel like I’m arguing about reality. Or maybe I am crazy and things like timestamps and multiple AI detectors (and just my own instincts that something is off) are giving me the wrong impression?

I wrote back and asked, “How’s his work aside from this? I’m skeptical you should be keeping him on and wondering what the rest of the picture looks like.”

Honestly, I think he struggles with time/task management. When I set really strict deadlines and help him break down the work into small chunks (outline by this date, draft by this date, etc.), the work is good. My armchair psychologist take is that he gets in a shame/insecurity spiral when he feels like he’s falling behind, leading to shortcuts and dishonesty. But whenever I try to get to the underlying problem, he says the technology is lying, and I’m not sure how to dig deeper. How do I correct a missed deadline if he insists (to the point of getting emotional) that the deadline wasn’t missed?

I think you’re probably going to end up needing to fire this guy, but let’s break it down a little more.

First: is it acceptable for the person in Miles’ position to need that much hand-holding with time management? Is it okay that you have to break work into such small chunks for him, or does the person in this job need to be able to manage their own work more independently? If it’s the latter, it might make sense to just focus there: he needs to manage the work more autonomously, without missing deadlines, period. If he can’t do that, there’s your answer.

But for some jobs, it would be more reasonable to help him with those aspects of the job. If that’s the case here, then let’s look more closely at the defensiveness, denials, and potential lying.

The first thing to do there is to bring in IT. If Miles says that the timestamps on shared documentation are wrong or otherwise claims technology is malfunctioning, then let’s take that very seriously and ask IT to investigate. Most likely they’ll confirm that he’s wrong, but who knows — it’s worth verifying with them, at least. From there, if Miles continues to blame technology, you can hold firm: tell him that IT says he’s wrong, that these supposed tech snafus aren’t happening to anyone else, and you need him to meet deadlines, period.

You should also make a point for a while of checking right when something is due to see if he has submitted it or not — if he hasn’t, you can speak to him about it on the spot and sidestep the technology claims entirely.

With the AI concerns: focus less on whether he’s using AI and more on the qualities of his work that are making you suspect it. There’s something about his work that’s ringing those alarm bells for you — stilted writing, overly buzzwordy, weird leaps in logic, whatever it is. Focus there; those things are work quality problems and they’re easier to address without getting drawn into a debate about whether he did or didn’t use AI. (Although if he is using AI, he’s probably leaving tracks on his company computer that your IT team could find for you, if you want to know for sure.)

It’s possible that by approaching it this way, it will change things for the better — maybe there really have been some tech problems and by getting them fixed, it’ll let you both move forward without that distraction, or maybe by focusing on the specific weaknesses in his writing he’ll be able to use that feedback to improve. I’m guessing not — but the advantage of going into it with an open mind is that you’ll see pretty quickly that either he can do the job at the level you need or not, and you won’t need to get bogged down in the framing he’s currently trying to use (which is “I’m not responsible for any of this”).

You may also like:
{ 124 comments }

{ 124 comments… read them below or add one }

  1. Sola Lingua Bona Lingua Mortua Est*

    I think you’re probably going to end up needing to fire this guy, but let’s break it down a little more.

    100%. Shifts blame, makes demonstrably false claims, isn’t accountable on either quality or timeliness? Take a long, hard look at why you want to keep him–is it just to avoid recruiting his replacement?

    Reply ↓
      1. duinath*

        Same. Adding the frustration and “visible emotion” when he’s called out and needing handholding to get things done is just salt on the wound.

        Reply ↓
      2. Caramel & Cheddar*

        Yeah, even without the “breaking company policy” part, I was confused how he was still working there or wasn’t at least on a PIP.

        Reply ↓
    2. Person from the Resume*

      Very much agreed. He’s derailing valid feedback and consequences by getting “defensive, frustrated, and visibly emotional.”

      I’d go with “software timestamps do not lie. Please stop denying you were late.” (Perhaps have IT investigate if he’s the only one using the software, but if it works for everyone else but him … don’t even bother IT.)

      Checking at the time it’s due is great advise too.

      LW needs to work on staying unemotional and keeping on track. And also asking him to stop denying, getting frustrated, and emotional.

      I know this isn’t easy, firing this guy might be easier. I might go to that because he’s not a good worker and you believe he’s a liar. Plus he has have two recurring problems where he’s claiming the software is lying instead of him.

      Reply ↓
    3. Peanut Hamper*

      Notice that OP wrote “dishonest but gets emotional…” (emphasis mine).

      That shouldn’t be a “but” — it really should be an “and”. The latter doesn’t excuse the former in any way.

      I do believe Miles’s time here is limited.

      Reply ↓
    4. OhGee*

      for real. “gets defensive when caught in lies, doubles down with more lies” is not something that will change easily.

      Reply ↓
  2. Keymaster of Gozer (she/her)*

    Consult with IT. Not only do we have ways of finding out if someone is telling the truth but it’s also a very common request! I’ve lost count of the number of times I’ve sent back system logs and other proof that someone was/wasn’t logged on when they said they were or deleted a file they swear they didn’t.

    We can even show sometimes where the information was copied from.

    Faced with the proof from IT it generally goes one of two ways:

    1. Person admits they panicked and asks for leniency and HR handles it from there
    2. Person swears that it’s a conspiracy and IT have edited the information. HR rolls their eyes and puts them on a disciplinary.

    Although, being *from* IT I’m very biased against anyone who tries to blame the technology for their own faults/laziness.

    Reply ↓
    1. N*

      I really appreciate your IT perspective here. I often feel like I’m bothering IT when I go to them with something like this

      Reply ↓
      1. Djs*

        As an IT person, I would *want* to know if there was a potential problem with something under my domain being wrong. Even if the answer is always “yup, I checked it, this person is lying”, the last thing I want is for people to lose trust in whatever I’m providing for them. Even if the loss of trust isn’t based in facts.

        Reply ↓
      2. Keymaster of Gozer (She/Her)*

        Trust me, you’re not. We love to pretend we’re Sherlock Holmes on the holodeck given half a chance :)

        Reply ↓
        1. Djs*

          You actually want to make them come to you, not “damnit- that weirdo is doing that crappy Patrick Stewart accent again” :)

          Reply ↓
    2. I'm just here for the cats!!*

      This is great advice! I’m also wondering if IT could block AI websites. I’m actually surprised it hasn’t been blocked since it doesn’t seem to be something the company wants people to use.

      Reply ↓
      1. Impending Heat Dome*

        Could be part of a program that’s part of their office suite. Photoshop has generative AI in it, for example, and CoPilot has been around a bit too.

        Reply ↓
      2. ProdMgr Replaced By AI*

        You can block AI websites but it only gets you so far. You can’t stop someone from using AI on a non-company device and then copying it over, for example.

        Reply ↓
        1. Observer*

          You can’t stop someone from using AI on a non-company device and then copying it over, for example.

          Well, to some extent, you can. Not with the snap of a finger, to be sure, but it’s definitely possible to make it harder.

          Also, if that’s what he’s doing, it’s generally possible for IT to figure this out. And it would be such an egregious over-step, because it means really thought out breaching of company policy, that it would make the real problem clear.

          Reply ↓
          1. Serena*

            Stopping someone from doing something isn’t really a “to some extent” thing. You either can stop them, or you can’t stop them. And here, they’re right that you can’t stop them.

            Reply ↓
        2. Snudence Prooter*

          You can tell them that if they’re doing company work, it has to be on company computers, though, especially when you’re trying to short circuit the suggestion of AI generated work. Plus, that means easier access to the time stamps on the documents, which would help him show he’s completing the work on time. So he should bow out as a double win!

          Reply ↓
          1. Snudence Prooter*

            You can tell them that if they’re doing company work, it has to be on company computers, though, especially when you’re trying to short circuit the suggestion of AI generated work. Plus, that means easier access to the time stamps on the documents, which would help him show he’s completing the work on time. So he should count it as a double win and definitely love the idea, as someone who is completely innocent!

            Reply ↓
      3. Sneaky Squirrel*

        Even if they could, would that a hinderance to advancement for IT to do so though? Generative AI seems like it’s going to be here to stay and it’s likely only going to get more robust. I think it would be smarter for companies to adapt training and policies around AI rather than shielding their staff from it.

        Reply ↓
    3. Ansteve*

      I’m in IT. I will go to bat for a user with genuine issues that stop their work. But if your manager calls the helpdesk asking if a system has been down all morning that we show was up and no reports of issues were made, I am not responsible for that bus I just threw you under.

      Reply ↓
      1. Lana Kane*

        Love that analogy but I see it more like being a spectator to that gorgeous swan dive they just took in front of the bus.

        Reply ↓
  3. Caramel & Cheddar*

    “There’s something about his work that’s ringing those alarm bells for you — stilted writing, overly buzzwordy, weird leaps in logic, whatever it is. Focus there; those things are work quality problems and they’re easier to address without getting drawn into a debate about whether he did or didn’t use AI. ”

    I think both are necessary here — LW says that it’s against their org policy to use AI for the tasks he’s been assigned. Providing feedback about the quality is important, obviously, which he would need regardless of how he’s coming up with the content. But since AI isn’t allowed for these documents, a zhuzhed up version of his AI first draft is still breaking company policy, which seems like something worth addressing.

    Reply ↓
      1. Caramel & Cheddar*

        Sure, though as Alison mentioned, there are probably traces on his computer that IT can find (e.g. browser history, etc.).

        Reply ↓
      2. Caroline*

        If they use Microsoft products, there’s often a version history in the document that will be very different if the document was written from scratch vs copied and pasted in bulk from AI and then edited a bit.

        Reply ↓
    2. NetClari*

      I agree with this. in some industries, the data privacy concerns from using AI would be a much bigger concern than the actual work product. I like Allison’s solution of asking IT to help check if he is accessing AI tools while writing. I agree that AI checkers don’t do anything.

      Reply ↓
      1. Cat Tree*

        Yes. My company is pro-AI and had even developed their own generative AI tools. BUT for data privacy we can’t use any outside AI and we have strict boundaries around certain things that can’t even use our company’s AI.

        Reply ↓
      2. ferrina*

        Yeah, if he’s putting proprietary information into a public AI engine, that data will now go to the AI company. That’s a big issue. IT should be able to help track if he’s accessing AI tools on his work computer.
        And if he is, that could help speed the path to his dismissal (which given all the other issues, seems inevitable)

        Reply ↓
  4. Madame Desmortes*

    Depending on the kind of place you work in, you could also move for some technological policy enforcement — ask IT to block chatbots on the work network. If that tanks Miles’s output, well, you know something about how he works.

    Reply ↓
    1. Bananapants Modiste*

      If I were Miles using AI to do my work, blocking chatbots at work would make me take my work home and have the AI write it there.
      But maybe there’s a policy about that, too.

      Reply ↓
      1. Strive to Excel*

        At that point you put in a computer-specific block & require all internet connections to be done either through a company VPN or through the company’s own internet. Those are useful restrictions to have on company computers anyways.

        Reply ↓
      2. Lenora Rose*

        A true thing about cheating and crime is that a lot of what we call “preventative” measures can’t actually STOP the cheating or crime as long as someone is determined. What they do is make it inconvenient, make it take time, make them require extra steps to accomplish the goal.

        And 90% of the time or more, it’s enough. A lot of people are looking for the easy way, and putting up blockades makes them go, “oh, it’s too much work now” and stop. And of the remaining batch, for another chunk of cases, forcing them to take more complicated steps makes them do something that exposes them.

        So talking about how “Well if you do this, they’ll just get around it by…” is actually not always true.

        Reply ↓
        2. Insert Clever Name Here*

          Yeah, exactly. It’s why locking your car door is such an effective way to keep it from being broken into. Someone can absolutely still smash your windshield, but what’s overwhelmingly more likely is that they’ll just go on to another car in hopes that it’s unlocked.

          Reply ↓
        3. bamcheeks*

          Yes! Plus many people are very good at convincing themselves that even if you know something’s not ~~technically~~ allowed, if it’s easy to do, it probably isn’t that big a deal. The more defences you have to work through, the harder it is to trick yourself that it’s OK really.

          Reply ↓
        4. metadata minion*

          Yes — especially in this case, where it looks like this is someone who’s overwhelmed and panicky. That is, it’s very unlikely that he has some deep desire to use AI specifically; he’s turning to it because he needs to hand in something fast, and if it’s now really annoying to access, AI isn’t a quick cheat anymore.

          Reply ↓
        5. Bilateralrope*

          True. But you don’t have to tell everyone about the anti-cheating method immediately. Giving you time to catch the employee and fire them if their behaviour is bad enough.

          In this case, the employer also has the advantage of knowing how the work should be completed. So they could install monitoring software, wait for the employee to submit documents, then look to see if the logs show the documents were produced in an approved way. If they don’t, the employee has some explaining to do.

          Reply ↓
      3. Observer*

        take my work home and have the AI write it there.
        But maybe there’s a policy about that, too.

        There should be. And those policies can be enforceable.

        Reply ↓
    2. Leaving academia*

      The letter writer said AI is appropriate in other circumstances, but not for external-facing documents. Assuming they mean the company doesn’t have a blanket ban on generative AI, and assuming it doesn’t have own internal system (or copilot or gemini), then other employees could be affected by a block when not doing anything against policy.

      Reply ↓
  5. Judge Judy and Executioner*

    If he’s lying about stuff that is easily disproved like time stamps, what other lies is he telling? I don’t think I could look past blatant dishonesty in a direct report. I would be validating with IT that he’s lying, and then have a serious conversation with him. If he gets emotional, recommend the EAP, and let him know you will give him time to calm down, but you need to revisit the topic with him the next business day to resume the conversation. It sounds like he need therapy to work through his emotions when he gets caught in his lies, and this is beyond PIP territory. PIPs are for things like “needs to improve time management and work more autonomously”, not for employees that have problems with the truth.

    Reply ↓
    1. Observer*

      If he’s lying about stuff that is easily disproved like time stamps, what other lies is he telling? I don’t think I could look past blatant dishonesty in a direct report

      Very much this.

      Reply ↓
  6. Zahra*

    Not to go towards armchair diagnosis… his struggles are similar to those people with ADHD live with. Whether he has ADHD or not is beside the point. If the struggles are similar, it’s a good idea to see if the solutions work for them. It’s not the LW’s responsibility to point their employee towards ADHD resources, but if the employee was writing, I’d definitely suggest it.

    However, for the shame/insecurity, if you have an EAP, please do direct him that way!

    Reply ↓
    1. Mentally Spicy*

      You can’t diagnose something like that from a letter written about a third person. Not should you. Whether he has some undiagnosed or undisclosed condition or not I don’t think it materially changes the advice anyway.

      Reply ↓
      1. Strive to Excel*

        I believe Zahra is saying whether or not he actually has a condition, LW could try X solution anyways because the problem he’s having could be helped whether or not there’s a diagnosis involved.

        Reply ↓
        1. Anonym*

          Yeah, a lot of the time and task management advice and resources for ADHD are useful to the wider audience of folks who struggle with those things. Appreciate the point about not diagnosing, but I think Zahra’s within bounds here.

          If OP wanted to give the advice, it might be framed as “time and task management” but that might lead them to ADHD resources either way.

          Reply ↓
    2. TinkerTailorSolderDye*

      Actually, I’m in agreement with Zahra here; not diagnosing, but wheeeeeeeew boy this sounds damn near identical to my stepson’s problem with the truth; a good, valid possibility, and I’m bringing it up to the psychiatrist this afternoon when we go in.

      Reply ↓
      1. Nobby Nobbs*

        I was just thinking that it sounds a lot like my problem with the truth during my own teenaged ADHD/anxiety/depression feedback loop era. In case it helps with your stepson: the problem wasn’t that I thought lying was okay, exactly. It was more that I’d gradually lost the ability to prioritize long-term consequences over short-term discomfort, and the lies became a (scarily easy, after the first few) means to push the immediate consequences of not doing something I should have just a little farther into the future.

        Reply ↓
        2. ferrina*

          Yeah, I’m ADHD and this pattern is not uncommon with ADHD people. You miss a deadline because of time blindness, then you make up some kind of excuse, even “oh, I definitely sent that”. I did that a little when I first started in an office job, though I stopped within a year. I do know someone that took this approach long term and carried it into his marriage and would regularly lie to his spouse over little things to get out of trouble (not that he was actually in trouble, but if he didn’t do his share of housework, he would lie to his spouse about it to avoid that conversation). It predictably ended in divorce, and I was always confused as to why he did this. Your explanation shines some light to his thought process.

          But agree with Zahra that regardless of whether it’s ADHD spurred or not, it’s not acceptable and this manager needs to take steps to let this person go. ADHD symptoms can contribute to this behavior, but this behavior is still a choice and can’t be allowed to continue. (And like Mango Freak says below, as an ADHD person, it’s my job to manage my symptoms and get accommodations if needed; it is absolutely not my manager’s job, and guessing a person’s diagnosis and trying to manage their healthcare decisions for them would be wildly inappropriate for a whole different list of reasons)

          Reply ↓
    3. Mango Freak*

      Nope, shut it down. If you need accommodations at work, ask for them.

      I have *severe* ADHD. It’s not my employers’ job to guess that I might need accommodations–it’s their job to listen and respond reasonably when I request them (and ideally to create a work environment that makes this easier or less necessary). It’s also their job to hold my work and behavior to a professional standard.

      I don’t go around lying about easily disprovable things, and I hate when others do.

      Reply ↓
  7. Sloanicota*

    I feel the way to address this is probably more annoying than OP wants to undertake, but once you have shown him how to break down the work tasks into deadlines, you’re going to have to check in – yourself, not through an automated tool – to see if he’s meeting those interim deadlines earlier in the process. Have him email you draft one or whatever, and follow up with him the minute you don’t have it in your inbox when you were supposed to, and then the same with task 2 etc. This is annoying. You shouldn’t have to do it. If you have to keep doing it more than once, you can follow the suggestion to let him go because he can’t handle the time management aspects himself.

    Reply ↓
    1. RagingADHD*

      Yes, it did make me wonder why LW is relying on timestamps to know whether the work was completed on time. If it was late, didn’t that impact anybody? Why do you need timestamps?

      If the person downstream (whether LW or otherwise) has the work when they actually get to it, then it was functionally on time, whether or not the timestamp matches whatever the putative deadline was supposed to be. Not to excuse Miles, but why is the deadline set in such a way that nobody actually notices?

      If it’s something like blog posts or social media that are supposed to go out on schedule, then the solution is to not have Miles be the one who hits “post.” Then you will know whether his stuff is late, because someone will see it.

      Reply ↓
    2. Mango Freak*

      Since this is a chronic problem I’m sure it plays out a lot of different ways and we can just trust LW.

      Maybe sometimes LW hears from another coworker that they don’t have the thing they need. Finance complains about a late expense report. LW is personally looking at a file that mysteriously updates after they bring it up to employee. Etc.

      Not much to be gained from picking this apart I don’t think.

      Reply ↓
  8. Mad Scientist*

    Sounds a lot like my coworker who constantly has an absurd amount of “IT issues” usually right before a deadline. He has deleted files en masse multiple times. Any file he touches for any reason gets corrupted or deleted. 9 times out of 10, these “IT issues” are easily avoidable and entirely his fault. It’s exhausting. I know shit happens, but when it’s happening constantly and only to you, it’s time to consider that you’re the common denominator…

    Thankfully, I just found out that this coworker is resigning! Hooray!

    Reply ↓
  9. Jam on Toast*

    It’s not the technology, it’s the outcome that matters. Lots of people wrote bad copy and missed deadlines before shared drives and GenAI. I think you need to identity the professional shortcomings: an inability to meet agreed-upon deadlines, failing to communicate potential delays in a timely fashion, poor work product, a pattern of deflecting accountability and failing to act on feedback from their manager.

    That they may be hiding behind technology to shirk their duties is only tangentially relevant here. Lay out the pattern and then reiterate to this team member what the standards or expectations for your team are. They have to show sustained improvement. If they don’t, dismissal may be your only option.

    Reply ↓
    1. murderbot*

      Well the AI is also an issue if the company has a policy against using it because they’re concerned about how the information being fed in will be used.

      Reply ↓
    2. ArtK*

      It’s about the technology, too. Many companies have clear anti-AI policies due to legitimate fears of privacy violations and the AI mining corporate secrets for “training.”

      Reply ↓
    3. Observer*

      It’s not the technology, it’s the outcome that matters

      That’s not necessarily the whole story, though. In There are some real concerns about AI that smart companies *have* to worry about.

      It’s not that the LW should not address the other issues. But if the company has a policy specifically about the use if AI, then they *have* to address that as well. Because I’d be willing to bet that the policy is not really about the quality of the writing.

      Reply ↓
  10. Killani S*

    Once employees start lying, there is no going back. It’s an undeniable sign of untrustworthiness, lack of respect, and poor judgment.

    Reply ↓
    1. Peanut Hamper*

      A lot of commenters on this site often recommend lying to get out of something or to avoid something, and it’s almost always a bad idea.

      Once you have a reputation as a liar, people are going to realize that you are unreliable at best. We have fables and fairy tales that address this.

      A single lie in the heat of the moment? Okay, that’s forgiveable. But a habit of lying? That’s where I draw the line.

      Reply ↓
      1. Pastor Petty Labelle*

        A lie, oh I have plans that can’t be moved when boss is asking you to work unreasonable hours? Ehh, not really a lie, my plan that cannot be changed is not working until midnight every night for a month.

        Lying about turning in work on time is a big deal.

        Reply ↓
        1. KitKat*

          Agreed. There’s a huge difference between white lies that smooth over social situations and lies about work product. I definitely do NOT see commenters here regularly advocating letter writers lie about work product!

          Reply ↓
      2. Dawn*

        There is only a moral imperative to be truthful when the person asking has the right to the information they’re asking for.

        Reply ↓
    2. Seventeen Apples*

      I disagree somewhat. While lying does show poor judgement and lack of trustworthiness, this may be a trauma response that, if the employee recognizes and addresses can change.

      An example: I had an employee who lied a few times. I sat him down and went through impacts of his lies. He had a whole sob story about an abusive prior boss. I told him that I’m so sorry he experienced that, pointed him to our EAP, and reminded him that trauma response or not, he had to tell the truth when doing his job, so he needed to figure himself out. I provided a few other resources and reminded him that mental health is health and if he needed a few days of sick leave to start addressing issues now, he can use that sick time. He did and put in a lot of work to address the why behind the lying. I don’t work there anymore, but he’s pretty great according to my former teammates.

      I also had an employee, same story, but he kept lying through his teeth, so obviously it depends on the person. But I don’t think anyone is irredeemable (obviously depending on situation, if he lied about murdering someone, that is a whole other can of worms).

      Reply ↓
      1. ferrina*

        It can be a trauma response, but that’s a reason not an excuse. You aren’t responsible for the trauma you endured, but you are responsible for how you manage your trauma symptoms (PTSD/cPTSD is a mental health condition).

        A certain level of honestly is necessary in the workplace and part of professional conduct, and if you can’t meet that, then it’s reasonable for your manager to take steps (regardless of whether or not you have trauma; it’s never on the manager to diagnose/manage trauma, for a long list of reasons). The first step should usually be verbal feedback, and it sounds like in your first example, the employee responded to that and made the necessary changes. I’m not sure if anyone is irredeemable, but I do know that someone who doesn’t want to change won’t change.

        Reply ↓
  11. Scottish Beanie*

    OP, I think you’re getting into the weeds about how he feels, when you need to strictly address the patterns of late documentation and shoddy work. He knows he’s been put on alert that he’s not meeting goals, so his response should be to actively participate in mitigating the situation, not just wait for you to notice it. He’s hoping you remain flustered at his emotional outbursts, in an attempt to distract you from the fact that he’s not doing his job. I’ve worked with many people like this. Once you see the pattern, it’s impossible to miss.

    Reply ↓
    1. Pastor Petty Labelle*

      This. OP you are so focused on getting to accept what the technology says you are missing the big picture. The guy is missing deadlines and not turning out work product in a way that you need.

      Stop trying to get him to accept that the technology says X. Simply state, the work was not turned in on time and what was turned in was unacceptable — what are you going to do to make sure this doesn’t happen again. Work through solutions with him. If he throws up his hands and says But I did turn it on time, you need to get serious with him. Tell him his job is on the line and he needs to meet deadlines with acceptable work product. Then follow through on consequences.

      This is not a court of law. You do not have to prove beyond a reasonable doubt that he turned in work late or used AI.

      Reply ↓
      1. el l*

        Yes, and all beautifully said.

        OP can’t lose sight of that core of the situation: “…the work was not turned in on time and what was turned in was unacceptable.” And: Despite repeated feedback, he’s not fixing it.

        Take AI completely out of these discussions if possible – it only gives you a means of how terrible work was done. It doesn’t tell you why his work product was turned in late and bad, and why this keeps happening. And in the end, even this doesn’t matter – he has to fundamentally be the one who does the mental labor to fix the situation and keep his job. Not you.

        Reply ↓
  12. many bells down*

    “You should also make a point for a while of checking right when something is due to see if he has submitted it or not — if he hasn’t, you can speak to him about it on the spot and sidestep the technology claims entirely.”

    Honestly, I don’t think this is going to work. He’ll still claim he submitted it and there must be some other technical issue. It’s just going to be “oh must not have synced right” or “maybe there’s an internet problem”.

    Reply ↓
    1. Not Tom, Just Petty*

      My interpretation of the plan was to contact him AS something is due, like quarter to the hour and ask for an ETA.
      “Will you be submitting that at 11 today or can you get it to me a few minutes earlier?”
      I know it sounds like a “gotcha” thing, but I agree with many bells down that anything after the time is going to be, “but I submitted it. It’s the system! It lost it and my final draft on my desktop.”

      Reply ↓
      1. Sloanicota*

        Yeah, a few times you might have to do the “show me the X draft” (on your computer, right now) at the time it is due. If you see there’s still comments in it or corrections haven’t been made, you can flag that he has officially missed the deadline for that stage. This is way more handholding that I would recommend longterm but if you want to get to the bottom of this, it might be worth doing once or twice.

        Reply ↓
      1. Caramel & Cheddar*

        Or, as they work through this period of extended oversight, cc/tag/ping/message the boss when the work is submitted. If he says “I submitted it on time but forgot to let you know!” as he inevitably will, that’s just another data point towards what to do with his long term employment possibilities.

        Reply ↓
      1. Sybil*

        I have the same problem with the high school students that I teach and I wouldn’t be too sure this mindset in the work world doesn’t come from what worked for them in school. “I did turn in [online assignment] to the correct place. It’s just the website that is not working!” Which is what they also tell their parents who then want to blame me for the zero.

        I just ask them to pull up the website and show me what they tried and how it wasn’t working. Sometimes, indeed, they are trying to submit it incorrectly, but more often they have to admit the assignment is not actually complete.

        Reply ↓
        1. Lexi Vipond*

          My (university) students are usually emailing me, so I sometimes end up in the odd situation of asking them to please send me a screenshot of something that isn’t there. But sometimes it really isn’t and they can send me a picture of the blank space, and that’s useful information!

          Reply ↓
          1. Hawkwind1980*

            One of the things I like about the new LMS our college adopted this academic year is that it sends students a submission receipt when they have actually sent something. If they claim an assignment was submitted and the instructor isn’t seeing it, we can ask them to forward the receipt and then sic the LMS’s IT on the problem.

            Reply ↓
      2. ferrina*

        That’s what I was picturing. Treat it like an IT issue; walk through the steps with him to see where the breakdown is occurring. Mentally set aside time to walk over to his desk, watch him send it, then walk back to your desk. Or if it has magically disappeared from his desktop, walk him over to IT.

        Reply ↓
  13. ClosingtheDoomLoop*

    I find it funny how everything seems to boil down to either quit the job or fire the employee these days as a solution. I don’t necessarily disagree with those but I’d love to see an analysis of the solutions suggested here.

    Reply ↓
    1. Caramel & Cheddar*

      I do wonder how much letting a situation play out too long before writing in impacts those kinds of stats, e.g. if you’re at the point where your boss/employee is making you write into an advice column, chances are higher than average that things are either approaching a place where they’re soon to be out of control or they’re already well out of control.

      There are definitely letters where people ask for advice in advance of something potentially becoming a program, but those are probably much rarer comparatively.

      Reply ↓
    2. Sola Lingua Bona Lingua Mortua Est*

      I find it funny how everything seems to boil down to either quit the job or fire the employee these days as a solution. I don’t necessarily disagree with those but I’d love to see an analysis of the solutions suggested here.

      People and jobs are largely fungible in the modern labor market.

      Reply ↓
    3. Nonsense*

      It’s the same reason why people suggest divorce in relationship columns – by the time they’re writing in about a problem, it’s basically too late to fix it easily. Plus, as we often see in the updates here and occasionally in the Aunt Agony columns, the problem they wrote in about was just the one they could name and there were a lot more issues under the surface.

      Also, far too many people have been stuck with a coworker who is constantly coddled by management while not doing their job.

      Reply ↓
      1. Richard Hershberger*

        Also, there is strong selection bias toward extreme cases. Most office muddle along more or less fine and there is nothing to write in about.

        Reply ↓
    4. SuprisinglyADHD*

      The common thread I’ve noticed is that if a problem can’t be solved by discussing the issue directly and collaborating on solutions/workarounds, then there often aren’t many other options besides “accept it as it is” or “either you or [problem person] should not be in this job”.
      We readers get a very curated list of letters, particularly the ones that are super-specific and delicate (eg: how do I help this employee whose house burned down without overstepping boundaries), the letters where simple troubleshooting has already failed, or the batshit bananapants situations that make everyone go “wat.”

      Reply ↓
    5. Leaving academia*

      I think there is a decent amount of “try communication” and “decide if you are willing to put up with it,” but letter writers might be better about including what they have already tried or whether they are willing to put up with the situation? Could certainly be interesting to see if that has changed (both on it’s own, but also in relation to answers). Particularly related to quitting, I wonder if there is a cultural difference in “just deal with it”.

      Reply ↓
    6. bamcheeks*

      I sometimes see complaints on here that there is too much, “time to find a new job” and that it’s the AAM equivalent of DTMFA. I think both of those things are actually really clear examples of “be clear in what the limits of your power are, and don’t forget them”.

      Overwhelmingly, problems are either, “Something is doing Thing I Don’t Like, and I haven’t told them because [reason]” or “I have told them, but they are doing the thing I don’t like anyway”. And people really often get stuck at that stage — sometimes because they think they’ve told them, or the other person Should Just Know, or because they’ve told the other person but the other person doesn’t care or doesn’t understand or doesn’t have the power to stop, but they feel sure that there must be something they can do to make them care / understand / have the ability. That’s when they write to problem pages.

      If you’re a manager, you actually can enforce consequences, starting with flagging the seriousness, moving to PIPs and disciplinaries, ending with firing. But for lots of other relationships, especially where you want something to change at work but those people are superior to you, or problems with partners or husbands and wives, you don’t have the power to make someone. “Start looking for a new job” or “break up with him” are a realistic reminder of what you CAN change, and what is in your power. You might decide that’s not the right solution and choose to stay with them and tolerate the behaviour, and you might feel you have very little active “choice” if there are major barriers to doing those things. But IMO it is important to recognise that those are the parts which are in your power, and “make this person stop doing the thing” probably isn’t.

      Reply ↓
    7. nnn*

      What are you basing that on? I just looked through the last 2 weeks of letters on the site and that wasn’t the advice for most of them.

      Reply ↓
    8. Dust Bunny*

      I read old advice columns and, honestly, nothing has changed much. Even in relationships the advice in the 1940s was “accept it or break up”. I imagine that if we had had AAM in 1950 the advice about handling incompetent and recalcitrant employees would be a lot of “fire them”.

      Reply ↓
  14. DivergentStitches*

    I would also allow Miles the opportunity to let me know if he has any medical condition that would require reasonable accommodations around the time management issues. I am neurodivergent and I struggle with that as well. If I hadn’t told my manager about my diagnosis and I was struggling and he had these sorts of conversations with me, I’d 100% tell him and ask for a reasonable accommodation.

    If he doesn’t have, or doesn’t say he has, a diagnosis, OP can guide him to the company’s EAP for assistance if needed.

    Just saying – once, I screwed up and I did cover it up dishonestly in a panic. I’m not proud of it. But again, neurodivergence.

    Reply ↓
    1. ferrina*

      You’re being very kind, but I think that’s too kind. I’m ADHD, and while many of these issues are things I’ve dealt with (yes, including lying to cover my time blindness), it’s also not the manager’s job to manage their employee’s health conditions. And these behaviors (whether or not they are caused by a neurodivergence) sound like a LOT. This isn’t a good employee, and if this behavior can’t change very quickly, this job may not be the fit for this person at this time. And that’s okay- keeping someone who is causing problems in the workplace often doesn’t end well long term. This person isn’t doing great work, isn’t meeting deadlines, requires a lot of time from the manager, and is regularly lying and getting emotional around basic work processes (i.e., getting feedback).

      You also can’t ask if an employee is neurodivergent. They either need to volunteer the information, or you have to operate as though they do not have a health condition. If you’re in the U.S., it’s a violation of the ADA to assume that an employee has a health condition. If you ask “have you been tested for a neurodivergence?”, that can open the door to an ADA violation lawsuit (no matter how well-meaning you are).

      Reply ↓
  16. Machine*

    There is nothing inherently wrong with using AI, only that AI is worse at writing than people. AI looks to fake for external documents, internal who cares, and external document quality matters. LW’s company is focusing on the fact that he did not write it; the problem is that the quality is bad. I play around with Chat GPT all the time, and its logic boards are impersonal and fail at times. That is OP’s problem, that the documents make the organization look unprofessional, and thus, his work quality must improve. I would recommend LW to use AI to bounce and generate ideas faster to better reach the deadline.

    Reply ↓
    1. Best Coke Ever*

      For my uses, ChatGPT is great at coming up with a draft which I’ll then need to tweak a bit. But it often pulls in things I wouldn’t have thought of. As long as you know the limitations I think it can be a great tool

      Reply ↓
    4. bamcheeks*

      internal who cares

      I mean– lots of people? I am sure there are lots of use-cases where it doesn’t matter (and there are lots of external ones too), but the idea that internal communications can all be shit because who cares is a wild take on the value of internal communication.

      If you’re not adding any value beyond what AI can do, what’s the point of you?

      Reply ↓
    5. ArtK*

      There’s tons wrong with using AI. Whatever it generates *is* based on plagiarism. The publicly available models have mostly (all?) been trained on data scraped from the internet; despite a common myth, just because it’s available on the net, it doesn’t mean that it’s copyright-free.

      For technical stuff, you have *maybe* a 50/50 chance that the response will be correct. Again, the common models have been trained on things like Reddit, without anyone curating the training data; that means that bad answers are very easy to show up, because the training data is awful.

      Reply ↓
      1. Best Coke Ever*

        “Whatever it generates *is* based on plagiarism”

        This is just incorrect. There are so many uses for AI that you are lumping everything in with one pitfall

        Reply ↓
    6. JFC*

      Yes, I feel like LW’s frustrations with Miles are spilling out into generalizing AI as being bad. It’s certainly imperfect but has helped me out in many situations. I would not send out an AI-generated document internally or externally without reviewing it myself first, but it’s a great tool to use as a starting point or if you are stuck.

      Re: the timestamp issue — is there any chance that Miles’ computer is showing a different time than yours? Times on devices can kind of be notoriously inaccurate. I’ve seen wacky times on otherwise perfectly usable computers before, and my old alarm system was always an hour behind no matter how many times I tried to correct it.

      Reply ↓
  17. Double A*

    As a teacher, I see a lot of students doing exactly this. I’m working to teach them time management skills and also think critically about AI, but it’s an uphill battle.

    When it comes to AI use, I also think it’s important to spell out the WHY of your policy. In my teaching, I have a policy where if I suspect AI but the student has otherwise achieved the task, then I grade it as if they wrote it themselves because they have used the tool more or less effectively, and my feedback will address areas where it’s less than effective. But it’s my job to teach.

    If you don’t want AI used, explain why. If it’s for data scraping reasons, then it’s a hard line. If it’s because it’s poor writing, then you don’t even need to accuse him of using AI. You just send it back with the feedback about why it’s unacceptable. If he’s using AI to panickedly meet deadlines, then turning in something subpar late is going to lead down the path to firing.

    Reply ↓
    1. Good Lord Ratty*

      It’s… plagiarism. Your hypothetical student who used AI to write a paper for them didn’t complete the assignment; they got someone (or something) to do the work in their stead. The point of a writing assignment is learning to write well. If you don’t do the work of writing, you aren’t learning to write. You can mark an assignment written by a human person other than the student (e.g., if they’d paid someone $50 to write their paper for them), but that doesn’t mean the student did the work.

      Why is it suddenly ok when it’s not a third party person being paid to help the student cheat, but instead it’s a technological solution? It’s still cheating. It’s still not doing the work.

      Reply ↓
      1. Double A*

        This is incorrect; plagiarism is passing off someone else’s ideas as your own. You can’t plagiarize AI because AI doesn’t have ideas. That’s why AI use doesn’t fall under plagiarism under any honor code in any institution. It’s incredibly tricky territory for schools.

        Look, if a student wants to interact with AI to prompt their thinking and they get some useful ideas about that, I don’t see how that’s any different than a class discussion or reading articles related to the topic. You then need to have the skills to synthesize those ideas, wherever they came from, and AI won’t allow you to fake those skills.

        If a student just plugs a prompt into AI and hits go (which a lot of them do), the answers will be wrong. The writing is bad. It can’t follow the instructions. I can tell it’s AI. It’s those borderline essays that I will take at face value. The writing is kind of stilted, but it has a logical flow. It has supportive evidence that connects with the thesis. I don’t really know how (or even iff) the student interacted with AI to achieve the outcome, but they have achieved the outcome of writing a coherent, well-supported essay. AI does not do that on its own without someone having the critical thinking skills to coach it correctly. If you have the skills to use AI to write an essay that meets my criteria, you have demonstrated the thinking skills I’m looking for, and I will grade you accordingly.

        I mean, AI is a tool. I not only encourage students to use spell check, I require it. They are in high school; my expectation is that they have the basics of spelling down, and therefore can use a tool that helps them monitor that skill. If they understand the fundamentals of writing, they might be able to use AI to help them. If they don’t, AI won’t save them.

        I will say I strongly discourage the use of AI for high school students because I don’t think they generally have developed the writing and thinking skills to be able to use AI in a supportive way. But I *explain* this to them, I don’t use say, “AI is not allowed.”

        Reply ↓
  18. Richard Hershberger*

    Breaking tasks down: Many many years ago I was a department manager in a WalMart. (This is not nearly as exalted as it sounds, by the way.) The department managing was mostly hands-on, but we would be assigned another employee to cover the later shifts, after we were gone for the day, with a couple of hours of overlap in our schedules. There was this one guy who the others didn’t want, because he was not smart. It is hard to express this without it coming across as an insult, but think Forrest Gump. He was also a hard worker and a really nice guy. I lobbied to have him in my department, because I had figured out how to instruct him. If I simply told him the desired end state he would freeze up, no matter how routine the path was to achieve that end state. So instead I broke the task down into small steps: Take this stuff off the shelf and put it on a cart and take it to the back and put it on this shelf. Then take this other stuff from a different place in the back and put it on the cart, bring it out front, and put it on the shelf formerly occupied by the other stuff. Once I figured this out I could go home confident that the work would get done, where other employees who had no problem with how to do it would somehow find reasons not to. I loved the guy.

    For that role, this level of management makes sense. The role in the case of this letter appears to bear little resemblance. This guy, for whatever reason, simply sucks at it.

    As for using AI, I personally am pretty lenient, in that if the work product is good I don’t care how it was arrived at, within broad limits. The problem with using AI is that the output is poor. I can see someone successfully using it as a first draft or to break writer’s block, but in practice the people who use it surreptitiously rarely have the skills to turn it into something actually good.

    Reply ↓
    1. Not Tom, Just Petty*

      I think there is another a big factor in the difference between your employee at WalMart and OP’s employee in Spacely Sprockets: the guy. Your guy did not tell you that the boxes he needed to stock were still on the truck during his shift. Your guy didn’t tell you that the list of items to stock was missing so he had nothing to reference. Your guy didn’t tell you that he did stock everything, but someone else moved all the items back to the store room.

      Reply ↓
      1. CommanderBanana*

        ^^ This. This commenter definitely deserves kudos for learning how to manage an employee effectively, but this is apples to oranges.

        Reply ↓
    2. A Significant Tree*

      I read that as the point – some employees thrive with an adjustment to how they are managed, and some will not improve no matter what you try. OP’s employee is definitely in the latter camp because he’s prone to blaming and reacting, so it’s almost certainly not worth the effort to find the magic method of management.

      Reply ↓
  19. CubeFarmer*

    Miles sounds like the type of person who has always been able to get away with stuff like this, and he’s bewildered as to why this is no longer working.

    I would also follow up with IT to check on the time stamp issues (even though, yeah, you know that he’s lying.) The AI stuff is trickier, because the detectors are often wrong (I have run my own writing through them and it’s sometimes come back as AI generated.)

    One thing you could do is to ask Miles to send you daily updates/drafts about the work he is doing, and or, check in just before deadline as to the status of the project (i.e. look at the project.)

    Reply ↓
    1. Sloanicota*

      Yeah this is basically a form of weaponized incompetence, assuming that blaming tech/being bad at tech (such a fuzzy line!) is part of the shtick.

      Reply ↓
  20. CommanderBanana*

    Personally, I find the best way to handle people who always have mysterious “IT issues” is to take it really, really seriously, get IT involved, document everything, and act as though you really do believe them. It’ll usually become apparent very quickly that they are lying.

    Reply ↓
    1. FashionablyEvil*

      Right? If the company’s systems are actually eating documents or randomly moving things, that’s a serious problem! (My company had serious issues with Excel files on Teams a couple years back, for example.)

      Reply ↓
  21. Observer*

    Focus there; those things are work quality problems and they’re easier to address without getting drawn into a debate about whether he did or didn’t use AI. (Although if he is using AI, he’s probably leaving tracks on his company computer that your IT team could find for you, if you want to know for sure.)

    Yes. There is very little doubt you could find out for sure, without the need for AI detectors. And if you can’t you probably need to give a good hard look at your IT and information security policies and procedures.

    Reply ↓
  22. Birb*

    I don’t use AI detectors because they are unreliable, and because neurodivergent people tend to get labeled AI (I’ve uploaded decades old papers I wrote and they’ve been flagged as substantially AI) but I DO check each student’s version history to see if they likely used AI. It will show that they wrote a 1000 word paper in less than 5 minutes without backspacing or editing anything, no typos, and they’ll INSIST the technology is broken. I can’t imagine having to put up with that from an adult in a professional environment.

    Reply ↓
  23. el l*

    Noticed that OP didn’t answer the question when asked “How’s his work?” and instead speculated as to how he was feeling and why he was behaving how he was.

    It’s not your job to figure out his deal. What is your job is to get work done at the quality you need and at the time you need. Neither of which – given the ringing silence about his work quality – he’s meeting.

    Think you’re already in firing territory. All you’re doing now is documenting it.

    Reply ↓
  24. Cat Lady in the Mountains*

    So, I’ve been on the other side of IT issues no one believed until our IT team got involved. I had emails from an important client start not reaching me at random. They had time-stamped proof that they had sent them. I was about to be put on a PIP over it before my manager agreed to pull in IT (I couldn’t do this myself because of our IT escalation processes), and they discovered that system-level spam filters had started catching some emails from the client’s sender domain, and I would have had no way to receive or even know I was missing the emails. Turns out a temporary employee at the IT firm had adjusted the system-level spam filters without anyone’s knowledge.

    All that is to say – the open-minded approach but doing some serious digging really is the fairest way to go. If you learn he was lying, now you never have to wonder/doubt yourself for moving him out of the role. If it turns out by some crazy fluke that he was telling the truth, now you can fix the problem.

    Reply ↓

Leave a Comment

Before you comment: Please be kind, stay on-topic, and follow the site's commenting rules.
You can report an ad, tech, or typo issue here.

Subscribe to all comments on this post by RSS