You posted a job. You got 200 resumes. Now you're staring at them thinking: "How do I review these fairly? Can I just filter out people without degrees? And what do I do with the 195 people I don't hire?"

Most founders think: "I'll skim resumes for keywords, auto-reject anyone missing requirements, and delete the rejects when we're done hiring." What they don't realize: how you handle resumes determines whether you face discrimination lawsuits, violate federal record retention laws, or accidentally filter out your best candidates because of unconscious bias.

Founders who auto-reject based on degree requirements face disparate impact discrimination claims. Founders who delete rejected resumes before the 1-year minimum retention period face EEOC penalties and adverse inference in lawsuits. Founders who don't address hiring bias build homogenous teams that underperform and struggle to scale.

Here's how to review resumes fairly, what you can and can't automatically reject, and how long you must keep every application (yes, even the bad ones).

Should You Do Blind Resume Review?

Blind resume review = removing identifying information (name, schools, addresses, graduation dates) before evaluating candidates

The Case For Blind Review

The research is clear:

Resumes with "white-sounding" names got nearly 50 percent more callbacks than those with "black-sounding" names

A 2014 study found that managers of both sexes were twice as likely to hire a man as a woman

Using conventional résumé screening, about a fifth of applicants who were not white, male, able-bodied people from elite schools made it to a first-round interview. Using blind auditions, 60 percent did

What gets removed in blind review:

Personal identifying information is redacted or removed. This might include names, addresses, schools attended, and even dates of graduation

Why this matters:

  • Names signal gender, ethnicity, religion

  • Addresses signal socioeconomic status, race (via neighborhood)

  • Schools signal class, network, perceived prestige

  • Graduation dates signal age

Without this info, you focus on: Skills, experience, achievements, work samples

How to Implement Blind Review (Without Expensive Software)

Example - 15-person startup, California:

"We decided to ask job applicants to not include their name, address, college name, or graduation date on their resume"

Then: After receiving resumes for a position, assign each candidate a number and refer to applicants by their number until people are brought in for interviews

Low-tech option: Use a Sharpie to redact names, schools, addresses before printing/reviewing

Mid-tech option: Copy resume text into spreadsheet, filter out identifying info, assign candidate numbers

High-tech option: Use software like Blendoor, GapJumpers, or ATS with built-in blind review

The Limitations of Blind Review

Critical caveat: While blind resume reviews are a good start, they're just that: a start. After you've picked out candidates using the blind review technique, you still have to meet them for an interview. And when you meet them, you'll learn all the demographic information you worked so hard to avoid in the first place

Blind review addresses bias in screening. It doesn't fix:

  • Interview bias

  • "Culture fit" bias

  • Like-me bias ("went to same school!")

  • Confirmation bias in final decisions

What to do: Combine blind resume review with structured interviews, standardized questions, and diverse interview panels

When Blind Review Makes Sense

Use blind review when:

  • You have high application volume (50+ resumes)

  • You're in competitive talent markets (tech, finance, consulting)

  • You want to improve diversity outcomes

  • You have time to implement the process

Skip blind review when:

  • You have under 20 applicants (manual bias check is faster)

  • You're hiring for roles where school/network genuinely matters (e.g., sales role selling to specific industry)

  • You're already using structured rubrics with objective scoring

Can You Automatically Reject Based on Criteria?

This is where many founders create legal risk without realizing it.

Disparate impact = when a neutral policy has a disproportionate negative effect on a protected class

An employer whose screening test or requirement has a disparate impact on a protected group must show that the requirement is job-related and consistent with business necessity

The landmark case: Griggs v. Duke Power (1971)

An employer required a high school diploma or passing scores on intelligence tests for certain jobs, which disproportionately excluded African American applicants based on longstanding educational inequalities

The Supreme Court ruled this violated Title VII, even though there was no intent to discriminate.

Degree Requirements: High Risk

The problem with automatically rejecting people without degrees:

Degree requirements often do have a disparate impact against African American and Latino applicants

To defend a degree requirement, you must prove:

  1. The degree is truly necessary to perform the job (not just "preferred")

  2. There's no less discriminatory alternative that would work as well

Degree requirements are slippery, in part because it isn't clear exactly what particular skills, aptitudes, or abilities a degree confers (unless a particular degree is required for licensing, like a law degree or medical degree)

Example - 10-person startup, Texas:

High risk auto-rejection: "All candidates must have a 4-year college degree" (Disproportionately screens out Black and Latino candidates who could excel in the role)

Lower risk approach: "Bachelor's degree in Computer Science OR 4+ years professional software development experience OR completion of coding bootcamp + portfolio of projects"

Why this works: Provides alternative pathways to demonstrate competence without requiring a degree

What You CAN Auto-Filter

Lower-risk automatic screening criteria:

Required certifications/licenses (when legally required for the job)

  • CPA license for accounting role

  • Bar admission for attorney

  • Medical license for doctor

  • Real estate license for agent

Must-have technical skills (but test them, don't just rely on resume keywords)

  • "Proficiency in Python" → Give coding test to verify

  • "Fluent in Spanish" → Conduct portion of interview in Spanish

Geographic requirements (when genuinely necessary)

  • "Must be based in New York" → If role truly requires in-office work

  • "Must have work authorization" → Legal requirement

Years of experience (with caution)

  • "5+ years project management experience" → Generally acceptable

  • BUT: Could have disparate impact on younger workers (age discrimination)

Higher-risk automatic screening:

College degree (unless truly job-related)

Specific schools ("must be from Ivy League")

GPA requirements (can screen out people who worked through school)

Employment gaps (discriminates against caregivers, people with disabilities)

"Culture fit" algorithms (often proxies for "people like us")

The AI Complication

If you're using AI/algorithms to screen resumes:

The EEOC has emphasized that employers using software/algorithms/AI as "selection procedures" can face disparate impact liability if outcomes disproportionately exclude protected groups

Companies need to implement regular audits of their AI hiring tools to check for disparate impact across protected classes. Human oversight should be mandatory, no candidate should be rejected by an algorithm without human review

Example - 12-person startup, California, using AI resume screener:

AI tool automatically rejects candidates who:

  • Attended HBCUs (historically Black colleges)

  • Have gaps in employment

  • Lack specific keywords

Result: Disparate impact discrimination claim + potential AI hiring law violations

Better approach:

  • Use AI to rank/score, not automatically reject

  • Require human review before rejection

  • Audit AI outcomes by protected class quarterly

  • Test for disparate impact

What to Do With Rejected Resumes

Short answer: Keep them. For at least one year. Longer is better.

Federal Retention Requirements

Private employers must retain personnel and employment records for one year from the date of making the record or the personnel action involved, whichever occurs later

What this means:

  • Application received January 1, 2026

  • Hiring decision made February 15, 2026

  • Must keep until February 15, 2027

Employment statutes require covered employers to retain job candidates' records—even for the candidates who aren't hired—including applications, resumes, interview notes, assessment tests, reference checks, drug screens and background screens

Best Practice: Keep for 2 Years

Best practice is to retain all job applications and resumes – solicited or unsolicited – for a minimum of two years from the date of the hiring decision

Why 2 years?

  • Government contractors have 2-year requirement

  • Some state laws require 2 years

  • Gives you buffer if discrimination charge filed

  • ADEA (age discrimination) has longer retention for workers 40+

What to Keep

Everything related to the hiring decision:

Not only do you need to keep the resume or application, you must also retain supporting applicant documentation. This would include items such as: interview notes, assessment tests, reference checks, background checks, all related documents leading to a hiring or non-hiring decision, as well as the offer or rejection letter

For rejected candidates, keep:

  • Resume/application

  • Cover letter

  • Any test results or work samples

  • Interview notes (if they got to interview stage)

  • Rejection email

  • Reasons for rejection (documented)

If a Discrimination Charge Is Filed

When an EEOC charge has been filed against your company, you should retain personnel or employment records relating to the issues under investigation as a result of the charge until the final disposition of the charge or any lawsuit based on the charge

What this means: If someone files an EEOC complaint, you must keep ALL related records until the case is fully resolved (even if that takes years)

How to Store Rejected Resumes

Physical resumes:

  • Secure file cabinet or storage room

  • Locked to protect candidate privacy

  • Label clearly: "Job Title - Hire Date - Destroy After [Date]"

Digital resumes:

  • Secure folder structure

  • Password-protected

  • Access limited to HR/hiring managers

  • Backup regularly

ATS (Applicant Tracking System):

  • Most ATS platforms auto-retain for compliance periods

  • Set retention policy in system

  • Export data before deleting if you switch systems

When to Destroy Records

After retention period expires:

When you're ready to "toss" those applications with confidential data, shred it or safely delete the electronic data

Don't just throw in trash: Resumes contain PII (personally identifiable information) - name, address, phone, email, potentially SSN

Proper disposal:

  • Physical: Cross-cut shredder

  • Digital: Secure deletion (not just "delete")

  • Both: Document destruction with destruction log

State-Specific Considerations

California

  • California has extensive record retention requirements

  • Recommend 2-3 year retention minimum

  • Strong disparate impact protections under FEHA

Colorado

  • Follow 1-year federal minimum

  • Colorado AI Act (effective June 30, 2026) requires documentation of AI hiring decisions

  • Keep records of how AI tools were used in screening

New York

  • Follow 1-year federal minimum

  • NYC Local Law 144 requires bias audit records for AI tools

  • Strong state-level disparate impact law

Texas

  • Follow 1-year federal minimum

  • TRAIGA (effective Jan 1, 2026) requires AI governance documentation

  • Note: Disparate impact alone not sufficient for discrimination claim under TRAIGA

Florida

  • Follow 1-year federal minimum

  • No additional state-specific requirements beyond federal

Practical Workflow for Resume Management

Step 1: Receive Applications

Set up organized system:

  • Folder structure: [Job Title]/[Date Posted]/Applications

  • Assign candidate ID numbers for blind review (if using)

  • Log receipt date

Step 2: Initial Screening

If using blind review:

  • Redact identifying info

  • Assign numbers

  • Score based on skills/experience only

If not using blind review:

  • Use structured rubric to reduce bias

  • Document why each person advances or is rejected

  • Avoid vague reasons ("not a culture fit")

Be specific:

"Lacks required Python experience"

"Has 2 years experience, role requires 5+"

"Not a good fit"

"Bad vibes"

Step 3: Document Decisions

For each rejected candidate, note:

  • Date reviewed

  • Who reviewed

  • Specific reason for rejection

  • Keep with their application

Why this matters: If you face discrimination claim, you need to show legitimate, non-discriminatory reasons for rejections

Step 4: Retention

Immediately after hire decision:

  • Move all applications to "Retention" folder

  • Set calendar reminder for destruction date (1-2 years out)

  • Restrict access to authorized personnel only

Step 5: Destruction

When retention period expires:

  • Review for any pending litigation (if yes, keep longer)

  • Securely destroy per policy

  • Log destruction: "200 applications for Marketing Manager role posted 3/1/26 destroyed 3/1/27"

Why How You Handle Resumes Matters

Resume review and retention isn't just logistics; it's legal compliance and bias prevention.

Best practices:

On blind review:

  • Use it if you have high volume or want to improve diversity

  • Combine with structured interviews (blind review alone isn't enough)

  • Simple implementation: Ask candidates not to include identifying info + assign numbers

On automatic rejection:

  • Avoid auto-rejecting based on degree requirements (disparate impact risk)

  • Provide alternative pathways to demonstrate qualifications

  • If using AI screening, audit for bias regularly and require human review

On retention:

  • Keep ALL applications (including rejected) for minimum 1 year, ideally 2 years

  • Keep everything: resume, notes, tests, rejection email, reasons documented

  • If discrimination charge filed, keep until fully resolved

  • Securely destroy after retention period

Three actions this week:

  1. Review your current auto-reject criteria: Do you filter out anyone without a degree? That's high legal risk. Add alternative qualification pathways.

  2. Check your resume retention: Do you have resumes from the last year? If not, you're violating federal law. Set up retention system now.

  3. Document rejection reasons: For your next hire, write specific, job-related reasons for each rejection. "Not qualified" isn't enough. "Lacks required 5 years SQL experience" is defensible.

The goal isn't perfect hiring; it's fair hiring with defensible documentation.

Blind review reduces bias. Thoughtful screening criteria avoid disparate impact. Proper retention protects you legally.

Do all three, and you'll hire better while staying compliant.

This content is provided for informational purposes only and does not constitute legal advice; for guidance on your specific situation, please consult with an employment attorney licensed in your state.

Reply

Avatar

or to participate

Keep Reading