Updated August 2026.
Seventy-nine percent of hiring managers say candidates should disclose when they’ve used AI to help with their application materials. Only about one in eight job seekers who actually use AI ever does. That’s not a small gap — it’s the entire debate in two numbers. Neither side is exactly wrong. They’re answering two different questions, and only one of them is the question that actually decides whether you get hired.
AI-generated illustration
The Real Version
Most advice on this question flattens into one of two takes: “never disclose — it’s just a tool, nobody discloses spell-check,” or “always disclose — honesty is the best policy, and they can probably tell anyway.” Both treat disclosure as a real action you take on the page. It mostly isn’t. No major résumé builder or applicant tracking system has a field for it — there’s no checkbox, no standard line, nowhere natural on a one-page résumé to put “AI-assisted” without it reading strangely next to your job titles. So for most applications, “should you disclose” isn’t really a decision you get to make in writing; the decision that actually exists is what you do with the tool, not whether you confess to using it.
That doesn’t make the 79% figure meaningless — it’s a real, measured expectation from Resume Genius’s 2026 Hiring Insights Report, a survey of roughly 1,000 U.S. hiring managers. But it sits next to a genuine inconsistency on the hiring side itself. In Resume Genius’s separate 2026 Hiring Trends Report of 1,500 hiring managers, 58% say their company now uses AI to screen resumes — and heavily AI-generated application materials are simultaneously the #2 most-cited candidate red flag, at 49%, just behind job-hopping. A year earlier, suspected AI-generated content was the single factor most likely to be an outright dealbreaker: 20% said it would “definitely” stop them from hiring someone, ahead of job-hopping, typos, or poor formatting. Hiring managers are, in short, using AI on their end while penalizing visible AI on yours — a double standard worth knowing about before you decide how loudly to announce anything.
| Statement | Share | Source |
|---|---|---|
| Hiring managers who say candidates should disclose AI assistance | 79% | Resume Genius, 2026 Hiring Insights Report |
| Job seekers who say they always disclose their AI use | 12.8% | Gdoc.io/Pollfish survey, 2026 |
| Job seekers who say they’d never disclose it | 54.2% | Gdoc.io/Pollfish survey, 2026 |
In Plain Language
Here’s the plain version: you almost never need to write “AI-assisted” anywhere on your résumé itself. There’s nowhere for it to sit naturally, and volunteering it invites exactly the “heavy AI content” flag that ranks as hiring managers’ second-biggest red flag. What you actually need is for nothing on the page to be true only because AI made it up — no invented numbers, no borrowed phrasing that doesn’t describe your real work, no skill you don’t actually have. That’s the real, evidenced line, and it has nothing to do with whether you used a tool.
Hiring managers back this up with specifics: 80% say they can often spot an AI-written resume “at a glance,” and they point to concrete tells, not a vague hunch.
| Signal hiring managers notice | Share who cite it |
|---|---|
| Unnatural phrasing or tone | 51% |
| Repetitive or generic language | 44% |
| Vague or inflated descriptions | 41% |
| Buzzword-heavy writing | 41% |
| Perfect grammar with no variation | 39% |
| Distinctive formatting quirks (e.g. em dashes) | 32% |
Every one of those tells describes generic, unverifiable writing — not the mere fact that a tool was involved. That’s the useful part: it means the fix isn’t hiding your AI use, and it isn’t confessing to it either. It’s making sure what’s on the page still sounds like a specific person describing specific, real work.
A Concrete Example
Picture one bullet point, written two ways. Version one, generated wholesale from a vague prompt: “Leveraged AI-driven strategies to spearhead cross-functional synergies and drive impactful results.” That sentence hits four of the six tells above at once — buzzwords, vague description, no real number, generic phrasing that could belong to almost any job. Version two starts from the same real accomplishment but uses AI only to tighten it: “Cut weekly reporting time from a full day to 90 minutes by restructuring the team’s data-pull process, after testing and rejecting two automated shortcuts that introduced errors.” Same underlying tool use. Completely different result, because the second version is still describing something specific and true.
That distinction lines up with the strongest evidence available on whether AI actually helps here. A large randomized study of 480,948 job seekers, published as an NBER working paper, found that algorithmic writing assistance — grammar and wording suggestions applied to a candidate’s own draft, not a resume generated from scratch — raised job-offer rates by 7.8% and pay by roughly 8.4%. It’s an important result, but its scope matters: that study tested editing help in 2021, before mainstream generative AI existed, so it’s evidence that AI-as-editor helps, not that AI-as-author does. Harvard’s own career-services guidance draws the same line without needing the study: “Generative AI should not be the primary author… Use generated text as a suggested edit, not as a final product.”
AI-generated illustration
Why This Matters to You
Four things to actually do with this. First, check the application itself before assuming a universal rule applies — Harvard’s own guidance says plainly that “some organizations may have rules about whether AI can be used in preparing application materials,” and if a posting or portal asks the question directly, that instruction always overrides any general rule of thumb, including everything above. Second, never let AI add a skill or a claim you can’t back up — this is the actual danger zone: 53% of job seekers admit to adding or considering adding a skill to their resume they don’t have yet, and AI tools themselves are the single most commonly padded skill, ahead of Office software or design tools. Third, if someone asks you directly, in an interview or a follow-up conversation, answer honestly and specifically rather than dodging — a companion piece on this site walks through exactly how to build that answer if you want the full framework. Fourth, use AI the way the evidence says it actually helps: as an editor tightening your own real material, not as the author inventing new material for you.
One more thing worth remembering before this feels like a one-sided ask: employers aren’t fully transparent either. Only 35% of hiring managers say their own company always discloses when AI is used to evaluate your application, and one in five don’t disclose it at all. The asymmetry is real, but it isn’t a special moral failing on your side of the table — it’s a hiring process still figuring out its own norms, on both sides, at the same time.
Sources & References
- Resume Genius — “2026 Hiring Insights Report: ATS, AI, & Employer Expectations”
- Resume Genius — “2026 Hiring Trends Report: AI, Ghost Jobs, and the New Rules of Getting Hired”
- Resume Genius — “2026 AI Impact on Hiring Report: Hiring Teams Embrace AI, but Authenticity is a Growing Concern”
- Gdoc.io — “36% of Job Seekers Use AI in Their Search, and Half Never Tell Employers” (via GlobeNewswire, August 19, 2026)
- Resume Genius — “2026 Job Seeker Insights Report: Stress, Uncertainty, & AI Adoption”
- Harvard University, FAS Mignone Center for Career Success — “AI for Resumes and Cover Letters”
- NBER Working Paper 30886 — “Algorithmic Writing Assistance on Jobseekers’ Resumes Increases Hires” (Van Inwegen, Zhang, Bertoglio Louro, Munyikwa, Horton)