How Companies Screen Job Applications With AI (And How to Beat It)

Last updated: March 2026

Almost every large company uses technology to screen your job application before a human sees it. 97.8% of Fortune 500 companies use an applicant tracking system (Jobscan, 2025), and 43% of organisations worldwide now use AI for HR and recruiting, up from 26% in 2024 (DemandSage). But the way this technology works is widely misunderstood, and the myths around it cause graduates to either panic or over-optimise in the wrong direction.

Here is what actually happens to your application, what the data says about AI screening, and how to make sure your CV gets through.

The 75% Auto-Rejection Myth (Debunked)

You have probably read that "75% of resumes are automatically rejected by ATS before a human ever sees them." This statistic is everywhere. It is also not true.

An Enhancv study of 25 U.S. recruiters, published through HR.com in 2025, found that 92% of recruiters confirm their ATS does NOT auto-reject resumes based on formatting or content. Only 8% have any form of content-based auto-rejection configured. The original 75% statistic traces back to Preptel, a recruiting company that went out of business in 2013 with no disclosed methodology.

What ATS systems actually do:

• Sort and rank applications based on keyword relevance and knockout questions
• Parse your CV into structured data fields (name, education, experience, skills)
• Flag candidates who meet minimum requirements for recruiter review
• Store all applications for compliance and future reference

The system does not delete your application. It organises it. The real challenge is not surviving an AI gatekeeper. It is standing out in a pool of 250-500+ applications that the average corporate role receives.

How ATS Systems Actually Parse Your CV

Different ATS platforms handle CVs differently, and understanding the major systems helps you format yours correctly.

The common thread: simpler formatting parses better across all systems. Single-column layouts, standard section headings (Education, Experience, Skills), and clean text without graphics or tables are universally safe.

The file format question

Modern ATS platforms (Greenhouse, Lever, Workday, iCIMS) all parse text-based PDFs reliably. The critical distinction is text PDF versus image PDF (a scanned document), not PDF versus Word. That said, Word (.docx) remains the safest universal choice, especially for older systems like Taleo. Keep both formats ready and submit whichever the application system requests.

The Real Bottleneck: Keyword Filtering

If ATS does not auto-reject your CV, what does cause applications to disappear? Keywords.

99.7% of recruiters use keyword filters in their ATS to sort and prioritise applicants (Jobscan). When a recruiter opens their ATS dashboard, they typically search for specific skills, tools, certifications, or job titles. If your CV does not contain those terms, it simply does not appear in their filtered results.

The data is striking: candidates who include the exact job title on their resume are 10.6 times more likely to get an interview (Jobscan). Yet 54% of candidates do not tailor their resume to the job description, significantly lowering their chances.

How to optimise for keywords

1. Read the job description carefully. Identify the skills, tools, and qualifications mentioned most frequently.
2. Mirror the exact language. If the posting says "project management," use "project management," not "managed projects." ATS keyword matching is often literal.
3. Include both acronyms and full terms. Write "Search Engine Optimisation (SEO)" so the system catches both versions.
4. Target 65-75% keyword match with the job description. This is the recommended threshold for passing keyword filters (Career Launch Campus).
5. Use the Skills section strategically. Greenhouse and similar systems weight the Skills section heavily. List 8-12 relevant skills that match the posting.

AI Beyond the ATS: What Else Is Screening You

ATS is just one layer. AI is now embedded across the entire hiring pipeline.

72% of HR professionals used AI in 2025, up from 58% in 2024 (HeroHunt). The primary use cases:

• Resume screening (56% of recruiters say this is where AI provides the most advantage)
• Candidate matching (AI suggests candidates from the database for open roles)
• AI video interviews (21% of U.S. organisations use generative AI for initial interviews; HireVue alone conducted nearly 20 million assessments in Q1 2024)
• Chatbots for initial candidate engagement and scheduling

62% of employers reported rejecting AI-generated resumes that lacked personalisation (Insight Global, 2025). This is a growing concern: over half of job seekers now use AI to write cover letters, and hiring managers can often tell. The irony is that 99% of hiring managers surveyed use AI themselves in some capacity.

The takeaway: use AI as a tool to help draft and refine, but ensure every application feels personally tailored to the specific role and company.

What Gets Your CV Past the 11-Second Scan

Even after passing the ATS filters, your CV faces a brutal human bottleneck. An InterviewPal study of 4,289 resume reviews across 312 recruiters found the average initial scan is 11.2 seconds (up from 7.4 seconds in TheLadders' earlier eye-tracking study). The median total review time for candidates who pass this scan is 1 minute 34 seconds.

What recruiters look at in those 11 seconds:

1. Current job title and company (or university and degree for graduates)
2. Skills section (quick match against the role requirements)
3. Most recent experience bullets (looking for relevance and impact)
4. Overall layout (clean and professional vs. cluttered)

77% of hiring managers immediately reject resumes with typos or grammatical errors (Resume Genius, analysing 500,000 resumes). Job seekers with over 99% accurate spelling are 3x more likely to be hired.

For graduate-specific CV guidance, see our graduate CV guide, which covers how to structure your CV when you have limited work experience.

Common Formatting Mistakes That Break ATS Parsing

These are the most frequent formatting issues that cause ATS to misread or misfile your information:

Headers and footers

Placing your contact information in headers or footers causes ATS to miss it 25% of the time (Jobscan). Always put your name, email, phone, and LinkedIn URL in the main body of the document.

Tables and columns

Multi-column layouts cause content to be merged or read out of order in many systems. Workday and Taleo are particularly bad at handling them. Stick to a single-column layout.

Graphics, logos, and icons

Most ATS parsers ignore images entirely. Star ratings for skills, progress bars, headshot photos, and decorative icons are invisible to the system. Use plain text instead.

Creative section headings

Headings like "Where I Have Been" instead of "Experience" or "What I Bring" instead of "Skills" confuse parsers that look for standard sections. Use conventional headings: Experience, Education, Skills, Certifications.

Unusual file formats

Submit PDF or DOCX only. Pages, Google Docs links, and image files will not parse correctly in most systems.

Your ATS Optimisation Checklist

Before submitting your next application:

1. Single-column layout with standard section headings
2. Contact info in the body, not the header or footer
3. Job title from the posting included in your CV
4. 8-12 relevant keywords from the job description, used naturally
5. Both acronyms and full terms for technical skills
6. No graphics, tables, or decorative elements
7. PDF or DOCX format (text-based, not scanned)
8. Zero typos (proofread twice, then proofread again)
9. Tailored to each application (never send the same CV twice)
10. Under two pages for graduates (one page is ideal)

For guidance on writing compelling bullet points and structuring your experience, see our cover letter guide and personal statement guide.

The Bigger Picture: Why ATS Is Not Your Enemy

The ATS exists because employers are drowning in applications. When a single role attracts 250-500+ submissions, some form of sorting is inevitable. The system is not designed to reject you. It is designed to help recruiters find the best matches faster.

Understanding this reframes your approach. Instead of trying to "trick" the ATS, focus on being a genuinely strong match for the roles you apply to. Tailor your CV, use the right keywords, format cleanly, and let the system do what it was built to do: surface relevant candidates.

And when the application route does not work, there is always direct outreach. Our guide on how to find and email hiring managers directly covers how to bypass the application pipeline entirely.

FAQ

Do ATS systems automatically reject 75% of resumes?

No. This widely-cited statistic has been debunked. An Enhancv study of U.S. recruiters found that 92% confirm their ATS does NOT auto-reject resumes based on formatting or content. The original claim came from Preptel, a now-defunct company with no disclosed methodology. ATS systems sort and rank applications but rarely delete them automatically.

Should I submit my CV as a PDF or Word document?

Both work with modern ATS platforms. Text-based PDFs parse correctly on Greenhouse, Lever, Workday, and iCIMS. The only issue is scanned/image PDFs, which cannot be parsed. Word (.docx) remains the safest universal choice, especially for older systems like Taleo. When in doubt, submit whichever format the application system requests.

How do I know which keywords to include in my CV?

Read the job description and identify skills, tools, and qualifications mentioned more than once. Mirror the exact language used. Candidates who include the exact job title are 10.6x more likely to get an interview (Jobscan). Aim for a 65-75% keyword match with the posting, and include both acronyms and full terms for technical skills.

Can recruiters tell if my CV was written by AI?

62% of employers report rejecting AI-generated resumes that lack personalisation (Insight Global, 2025). Use AI to draft and refine, but always personalise the final version for each specific role. Generic, obviously templated applications are the main red flag, whether human or AI-written.

How long do recruiters actually spend reading my CV?

An InterviewPal study of 4,289 resume reviews found the average initial scan is 11.2 seconds. Candidates who pass this scan get a median total review of 1 minute 34 seconds. This means your CV must communicate your strongest qualifications within the first few seconds of reading, primarily through your headline, skills section, and most recent experience.