Do employers check cover letters and résumés for AI?

Updated October 31, 2026 · 6 min read

How many employers check for AI?

The numbers come from surveys by companies that sell hiring or résumé services, so we name the sponsor and the sample for each.

Read the 61% carefully: it covers any "AI use in the hiring process," including deepfakes and scripted interviews, not only AI detectors run on cover letters.

Do applicant tracking systems detect AI?

• We found no built-in detector. Greenhouse names "AI-generated resumes" as a problem, but its Real Talent suite lists fraud detection from "objective digital signals," a spam blocklist and CLEAR identity verification, not an AI-writing detector. We found no Workday, Lever or iCIMS page describing one either. Named "AI content" modules for those systems appear only on blogs that cite no source.
• Detection is sold as an add-on. Pangram markets cover-letter and résumé screening to recruiters, with an API that can "tag incoming applications with an 'AI Likelihood Score'" in an applicant tracking system, and recommends "using AI detection as a signal, not a verdict." Sapling sells a résumé detector and warns that sections "under about 500 characters" are more likely to be flagged.
• Hiring vendors advise caution. The recruiting platform Gem says "Standalone AI resume-detection tools should not serve as the primary basis for screening decisions." QuillBot's detector page says "Never make an academic or employment decision on a detector score alone."
• LinkedIn writes cover letters too. Some LinkedIn Premium subscribers get "cover letter assistance" that drafts a letter from their profile and the job description. For its Premium Apply Assistant, LinkedIn said "Draft creation and AI assistance are not shown to recruiters" (as reported by HR Dive, June 2026).

Can recruiters tell when a cover letter was written by AI?

• Not reliably by eye. TopResume asked hiring managers to say which of four résumés were written by AI, and 33.5% spotted the AI ones, even though its headline says managers can spot one "in under twenty seconds."
• Admissions officers did a little better. In a 2026 Harvard and NYU study, admissions staff told AI-written essays from human ones with an AUC of 0.70 (0.5 is chance, 1.0 is perfect): "meaningfully above chance though far below the accuracy achieved by commercial detectors."
• Detectors catch fully AI-written résumés. In a 2025 NBER study, GPTZero, Pangram and Originality.ai missed none of the fully AI-generated résumés, and GPTZero and Pangram flagged none of 192 human résumés at their best settings (Originality.ai 0.13%). An open-source RoBERTa detector flagged 69% of the human résumés. Text you wrote and then lightly edited with AI is a much less predictable case. What the accuracy studies show.

Does using AI hurt your chances?

• In admissions, it did. The Harvard and NYU study looked at nearly 7,500 applications to a public-policy master's program that prohibited AI. In the latest cycle, 56% of applicants submitted at least one essay that GPTZero flagged as mostly AI-written, and "Applicants submitting AI-written essays were admitted less often than comparable non-users." It's a preprint, about admissions essays rather than job applications.
• Cover letters lost weight as a signal. When a large online labor platform introduced an AI cover-letter tool, the link between how well a cover letter matched the job and the chance of a callback fell by 51%, and employers shifted toward other signals such as work history. Time spent editing the AI draft was linked to more hiring success (Cui, Dias and Ye, 2025 preprint).
• Tailoring stopped paying off. On Freelancer.com, employers paid a premium for customized applications before LLMs arrived, "but not after" (Galdin and Silbert, 2025 preprint).
• Applicants are adjusting. In Greenhouse's October 2026 survey, 39% of US job seekers had rewritten an application so it wouldn't be mistaken for AI, and another 21% had considered it.

What do employers' own AI rules say?

Do any laws cover AI screening of applications?

• New York City. Local Law 144 bars employers from using an "automated employment decision tool," meaning machine learning, statistical modeling, data analytics or AI that substantially assists or replaces hiring decisions, including screening, unless it had a bias audit within the past year and candidates were told. Enforcement began on 5 July 2023. The city's guidance we read doesn't mention AI-text detectors, so whether a detector score used to screen applications falls under the law is a legal question.
• European Union. The AI Act lists AI systems used "to analyse and filter job applications" as high-risk. After 2026 amendments, those obligations apply from 2 December 2027. Whether a general AI-text detector used on applications counts depends on its intended purpose.

This is a summary, not legal advice.

How can you use AI on a cover letter safely?

1. Check the employer's rules. Look on its careers page, as the UK Civil Service and Anthropic publish theirs.
2. Write the first draft yourself, with your real reasons and examples. That's Anthropic's advice to its own applicants.
3. Use AI to polish, not to invent. Tighten sentences and fix errors; never let it add experience you don't have.
4. Edit the result. In the cover-letter study, time spent editing AI drafts went with more hiring success.
5. Never hide text for AI screeners. Hidden prompts in résumés are among the deceptive uses Greenhouse's hiring managers say they catch.
6. Keep your drafts, and be ready to explain how you used AI if you're asked.
7. Check the final version if you want to see whether any passage reads like AI. Short texts give less reliable results, so treat it as a rough guide. What detection scores mean.

By the AI Checker Plus team. We research and draft these guides with help from AI tools, then check every claim against the sources listed above.