Can Hiring Managers Actually Tell When Your Cover Letter Was Written by AI?

Quick answer

Most hiring managers believe they can spot an AI-written cover letter. In Insight Global's 2025 AI in Hiring survey of 1,005 U.S. hiring managers, 88% said they can tell when a candidate used AI to help with applications, resumes, or cover letters. But no independent study has actually measured how accurate that gut instinct is — it's self-reported confidence, not tested detection accuracy. What the research has measured is something more useful: a large labor-economics study found that once AI cover-letter tools became common, the cover letter itself stopped predicting who got hired. The real risk of an AI-written cover letter isn't getting "caught." It's writing something so generic it no longer says anything about you at all.

What hiring managers are actually reporting

AI use in job applications is now the norm, not the exception. Resume Genius's 2026 AI Impact on Hiring Report, based on a Pollfish survey of 1,500 U.S. hiring managers fielded in June 2026, found that 87% of hiring managers use AI somewhere in recruitment, and 80% have encountered AI-assisted job applications firsthand.

Fifty-eight percent of recruiters report encountering AI-generated resumes or cover letters. Nearly half, 46 percent, have detected candidates using AI to respond to interview questions. Thirty-four percent have discovered AI-crafted LinkedIn or other social media profiles. Eighty-two percent of hiring professionals say they are personally concerned about the use of AI in job applications. Eighty-six percent agree that AI will make it harder to verify whether a submission accurately reflects a candidate’s true abilities.

That final statistic carries the most weight. Managers tasked with hiring say they lack assurance that AI use can be spotted readily. Their sentiment leans toward the opposite: they feel authenticity checks are growing tougher. According to Insight Global, however, the majority assert they are able to identify AI on their own.

"I can usually tell" and "I can prove it" aren't the same claim, and that gap matters. A hiring manager can spot an obviously templated letter stuffed with stock phrases without much effort. Catching a cover letter that started as an AI draft and then got carefully edited is a different problem entirely. We've covered this same gap before: confidence in spotting AI text tends to run well ahead of what's actually measurable, whether the text in question is a resume or a student essay.

The real story: the cover letter stopped working as a signal

The more interesting finding isn't from a survey at all. It's from a real field experiment, run on an actual hiring platform with real outcomes at stake.

Jingyi Cui, Gabriel Dias, and Justin Ye’s working paper investigates the impact of Freelancer.com’s April 2023 “AI Bid Writer” cover-letter tool. Their analysis covered more than five million submissions for over 100,000 jobs, detailing who accessed the tool, who used it, and how usage affected callback and hiring rates.

The letters got better. Workers with access to the AI tool wrote cover letters that were more closely tailored to the job posting, and using the tool raised the probability of getting a callback by several percentage points. But the letters stopped meaning anything. Before the tool existed, a well-tailored cover letter strongly predicted who got an interview and who got hired. After it rolled out, that predictive power collapsed — the correlation between cover-letter quality and getting a callback fell by 51%, and the correlation with actually landing the job fell by 79%. Employers didn't get better at catching AI text. They just stopped trusting the cover letter and leaned harder on a different signal instead: each worker's platform review score, based on past completed jobs, became more predictive of getting hired than it had been before.

Wharton economist Judd Kessler, writing in Knowledge at Wharton, argues this is the whole story of the cover letter right now. "Enter large language models like ChatGPT and anyone can write, in a matter of minutes, a deeply thoughtful cover letter that might have previously taken an impressive candidate hours," Kessler wrote. "This innovation has dramatically lowered the cost of the signal, substantially degrading its ability to showcase that a candidate is a good fit for a job." In an interview with Fortune, he put it more bluntly: "AI killed the cover letter." His point isn't that detectors will start catching AI cover letters. It's that once everyone can produce a polished, tailored letter in ninety seconds, a polished, tailored letter stops telling an employer anything useful about who's actually motivated or capable. It becomes what he calls "cheap talk" — a signal anyone can fake, so nobody can rely on it.

What actually makes a cover letter read as AI-written

Detection here isn't about running text through software. It comes down to what a tired human reader notices on a first pass, and the patterns look a lot like the ones we cover in AI-generated writing tells.

Four tells show up again and again. Enthusiasm with nothing to anchor it, like "excited to contribute my skills," a line that fits any job at any company. A structure that never bends: intro, qualifications, close, no digression, nothing that only makes sense from one particular person. Details with nothing real behind them, no named project or team or event, because AI can summarize a posting but can't reference what nobody told it. And competence that's a little too even, every sentence at the same polish, missing the small unevenness real writing tends to have.

This material does not establish that AI was used. Numerous meticulous writers produce spotless, elegant prose. In addition, a lot of AI-written letters are transformed through heavy revision into distinct texts. A weary reviewer may see this pattern as a series of cues, not as a final decision.

Should you still use AI to write your cover letter?

Using AI per se is not the danger; the issue lies in submitting a generic product. Researchers Cui, Dias, and Ye identified aspects that merit careful consideration. When the AI tool was used, the applicants who poured more effort into editing their drafts before submission enjoyed superior outcomes compared to those who sent nearly unaltered copies. The tool itself is an effective kickoff. Problems surface when the cover letter remains a machine’s rough draft rather than a polished document.

Before you submit anything AI-assisted, it's worth checking how it actually reads. Run your draft through our free AI content detector — not because a hiring manager is necessarily going to run the same check, but because a detector flags exactly the kind of flat, generic phrasing that a human reviewer also notices, just faster and more consistently than a tired skim at 4pm on a Friday. If large sections come back reading as heavily AI-patterned, that's your cue to add the one detail — a real product you used, a specific line from the job posting, a number you can defend in an interview — that no AI draft will invent on its own.

For hiring managers: what to rely on instead of a hunch

88% confidence, zero validation studies. That's not a foundation for rejecting someone on a hunch. What holds up instead: reference checks, real work samples, portfolio links, sharp follow-up questions in the room — Kessler's "hard to fake" signals, backed by the Freelancer.com data. Want something steadier than reading tone off a page? Our detector gives a text-level score, one input among several, never the whole decision. Same caution we'd give any school applies to a hiring desk too.

FAQ

Can a hiring manager prove a cover letter was written by AI? Not reliably. Self-reported confidence is high (88% in Insight Global's 2025 survey), but that's a belief, not a validated accuracy rate. No published study has tested hiring managers' detection accuracy against a known set of AI-written and human-written letters the way researchers have tested text-detection tools on essays.

Does using AI to write a cover letter hurt your chances? Not on its own. The Freelancer.com field study found AI-assisted applicants got more callbacks on average, and heavier editors of AI drafts did better than light editors. The risk is submitting something so generic it fails to distinguish you at all, not the fact that AI touched the draft.

If cover letters don't work as a signal anymore, why do companies still ask for them? Mostly inertia. Judd Kessler predicts they'll either fade out as employers realize they're not reading them, or stay required and get uniformly ignored. Some employers have already shifted weight toward work samples, references, and specific interview questions instead.

Is there a way to check if my own cover letter reads as too AI-generated before I submit it? Paste it into our demo first. It catches AI-typical phrasing sentence by sentence, the same flat language a sharp reviewer would flag on their own.