AI resume screening is software that helps an employer sort applications before or while people read them. It comes in three forms: an applicant tracking system that stores and ranks applications, rules and filters that can reject you automatically, and AI matching that compares your resume with criteria the hiring team sets.
A person usually makes the final decision, but an employer’s filters can remove an application before anyone reads it. You get through by meeting the stated must-haves, answering screening questions accurately, and stating your qualifications in terms both a parser and a recruiter can follow.
Key takeaways
- An applicant tracking system is a database and workflow tool. What it filters out depends on the rules each employer sets.
- Rejection can be automatic where an employer sets a rule: a knockout question, or a filter on criteria such as a credential or a gap in work history.
- AI matching tools rank or sort candidates against the employer’s criteria. At least one major vendor documents that its tool does not reject anyone and that humans decide.
- The claim that 75% of resumes are rejected by an ATS has no study behind it. Survey evidence does show that employers’ filters exclude many qualified candidates.
- In the United States, anti-discrimination law applies to AI hiring tools, and New York City requires a bias audit and notice to candidates.
01
How does AI resume screening work?
“AI screening” covers different mechanisms, and they fail in different ways.
| Layer | What it does | Can it reject you without a person? |
|---|---|---|
| Applicant tracking system | Stores your application, parses the resume into fields, and lets recruiters search and sort | Not by itself |
| Knockout questions and filters | Apply the employer’s rules to your answers and your record: a license, a location, a credential, a gap in work history | Yes, when the employer sets the rule |
| AI matching | Compares resumes with criteria such as skills, experience and job titles, then scores or sorts candidates | It depends on the product and the employer |
Vendor documentation shows how the last two work. Greenhouse, a widely used hiring system, describes an auto-reject rule that rejects an applicant automatically based on the answer to a question. Its example is a driving job that rejects applicants without a Class A commercial driver’s license.
For AI matching, the same vendor’s Talent Matching feature sorts candidates into categories such as strong, good, partial and limited according to how closely each resume aligns with criteria the hiring team sets. Greenhouse calls it assistive AI, not automated decision-making, and says it does not auto-reject or auto-advance any candidate: the results go to recruiters and hiring managers, and humans decide. Other products and employers may work differently. Video and assessment tools are covered in AI interviews, and the underlying system in what is an ATS?
02
Does a person still read your resume?
Often, but not always. It depends on the employer’s rules.
The best evidence comes from a 2020 survey of employers in the United States, the United Kingdom and Germany for a Harvard Business School and Accenture report. About two thirds of the employers surveyed, 63%, used a recruitment management system, and more than 90% of those used it to initially filter or rank candidates. Of the employers that did, 48% filtered middle-skills candidates on an employment gap of more than six months, and the report notes that a recruiter never sees an application screened out that way.
Past the filters, a recruiter with several hundred applications works from the top of a sorted list. Ranking sets the order, so a resume near the bottom may never be opened.
03
Is it true that 75% of resumes are rejected by an ATS?
No source supports that figure. It circulates without a study, a sample or a method attached.
That does not make the opposite true. In the Harvard survey, 88% of employers using a recruitment management system said it filters out, at least sometimes, high-skills candidates who could do the job but do not fit the exact criteria in the job description. For middle-skills candidates the figure was 94%. Automated filtering is real, and employers themselves say it removes people they would want.
The survey does contain a 75% figure, for something else: the share of U.S. employers surveyed that used a recruitment management system at all. It counts employers using software, not resumes rejected by it.
These filters act on criteria the employer chose, such as credentials, experience and gaps. They are not a hidden test of fonts or file types. You cannot format your way past a requirement you do not meet, but you can state plainly the ones you do.
04
What rules apply to AI hiring tools?
Federal law in the United States. The Equal Employment Opportunity Commission says the laws against employment discrimination apply to AI and other new technologies just as they apply to other employment practices. Its fact sheet gives programming a resume screener to reject certain people based on a protected characteristic as an example of intentional discrimination, and adds that a seemingly neutral practice is also illegal when it has an unjustifiable disparate impact.
New York City. Local Law 144 prohibits employers and employment agencies from using an automated employment decision tool unless it has had a bias audit within one year of its use, information about the audit is publicly available, and candidates have been given notice. Enforcement began on July 5, 2023. According to the city’s FAQ:
- The law covers tools that use machine learning, statistical modeling, data analytics or AI to substantially assist or replace discretionary decision-making, including screening.
- It applies when the job is located in an office in New York City at least part time, or is fully remote but tied to an office in the city.
- Candidates who live in New York City must be notified at least 10 business days before a tool is used. The notice must name the job qualifications or characteristics the tool will assess and explain how to request a reasonable accommodation.
- The law requires the audit and the notice. It does not require an employer to act on an audit’s results, although anti-discrimination laws still apply.
Complaints about a missing audit or notice go to the city’s Department of Consumer and Worker Protection. Other states and countries have their own rules.
05
How do you get through AI resume screening?
- Apply where you meet the must-haves. Knockout rules are binary. No wording gets you past a license, a location or a work authorization you do not have.
- Answer screening questions slowly. “Do you now or will you in the future require sponsorship?” is easy to get wrong in a hurry, and a wrong answer can be an instant rejection.
- Use the posting’s terms for skills you have. Put them inside the bullet that shows the work, as described in resume keywords. Hidden white text and pasted keyword lists are obvious the moment a person opens the file.
- Make the resume easy to parse. One column, standard headings and real text, following the rules for an ATS-friendly resume.
- State facts explicitly. Write out job titles, dates, tools, credentials and years of experience.
- Explain gaps briefly. Employers do filter on them. Wording is in employment gaps on a resume.
- Look for notices and options. Greenhouse’s documentation says candidates see an AI disclaimer on the job post and can follow a link to manage their automated processing preferences.
- Add a human route. A referral or a short message to the recruiter puts your name in front of a person whatever the sort order.
Hi [Name], I applied for the [job title] role on [date]. Could you tell me whether an automated tool is used to screen or rank applications for this role, and whether an alternative review or an accommodation is available? [If relevant: I am asking because (brief reason, such as a disability-related need).] Thank you, [Your name]
FAQ
Frequently asked questions
How does AI resume screening work?
Software stores your application, applies any knockout rules and filters the employer set, and may score or sort resumes against criteria such as skills, experience and job titles. Recruiters then work from the sorted list. People usually make the final decisions, but the software shapes which resumes they see.
Do applicant tracking systems automatically reject resumes?
They can, when an employer sets a rule. Greenhouse, for example, documents auto-reject rules for yes/no and multiple-choice screening questions, and employers surveyed for a Harvard Business School report said they filter on criteria such as gaps in work history. Without such a rule, an ATS stores and ranks applications for people to review.
Is it true that 75% of resumes are never seen by a human?
There is no traceable study behind that number. Automated filtering is real, though: in a Harvard Business School and Accenture survey, 88% of employers using recruitment software said it sometimes filters out qualified high-skills candidates who do not match the exact criteria. Treat any precise rejection percentage without a source as marketing.
Is AI resume screening legal?
In the United States, employers may use it, and the EEOC says anti-discrimination laws apply to AI just as to other employment practices. New York City adds requirements: a bias audit within a year of use and notice to candidates. Rules elsewhere differ. This is general information, not legal advice.
Can you opt out of AI resume screening?
Sometimes. It depends on the employer, the software and where the job is. Greenhouse’s documentation describes a link on the job post for managing automated processing preferences, and New York City’s required notice must explain how to request an accommodation. If you see neither, ask the recruiter what alternatives exist.
Sources
- NYC Department of Consumer and Worker Protection: Automated Employment Decision Tools: Local Law 144’s bias audit and notice rules and the enforcement date; the agency’s FAQ explains which tools, jobs and candidates are covered.
- U.S. Equal Employment Opportunity Commission: What is the EEOC’s role in AI?: how federal anti-discrimination laws apply to AI used in employment decisions.
- Greenhouse Support: Auto-reject: how an employer sets a rule that rejects applicants based on a screening answer.
- Greenhouse Support: Talent Matching FAQ: how an AI matching feature sorts candidates, who decides, and what candidates see.
- Harvard Business School: Hidden Workers: Untapped Talent: the employer survey on recruitment systems, filtering and ranking.
Published by Jobbie and last updated on October 6, 2026. This guide is general information for job seekers, not legal, tax or financial advice. Spotted something wrong? Tell us.