Blind hiring workflows aim to remove bias from recruitment. Pronouns are a common source of unconscious gender bias but often slip through manual anonymisation. This guide explains how to detect and handle pronouns effectively to keep your blind hiring process consistent.
TL;DR: Pronouns reveal gender and can undermine blind hiring. Manual removal is slow and error-prone. Regex scripts catch most common pronouns but miss edge cases. AI tools can identify pronouns in context but need training and oversight. Distill offers one-click pronoun removal that fits into your workflow without fuss.
Why pronouns matter in blind hiring
Pronouns like "he", "she", "they", or "ze" signal gender identity. Their presence in CVs or cover letters can trigger unconscious bias, even if names and photos are removed. Under blind hiring principles, all bias signals should be stripped.
The Equality Act 2010 doesn't specifically mandate removing pronouns. But agencies aiming for fair hiring often go beyond legal minimums to reduce bias. Pronouns are a subtle but real bias signal that affects decision-making.
Ignoring pronouns risks skewing candidate assessments and undermines the credibility of blind hiring. Recruiters report that pronouns often slip through because they're embedded in free text and harder to find than names or contact details.
Method 1: Manual removal
The simplest method is manually scanning CVs and cover letters for pronouns and replacing or deleting them.
Pros:
- Immediate, no technical setup
- Full control over what's removed
Cons:
- Time-consuming and prone to error
- Easy to miss less common or neopronouns
- Difficult to scale for large volumes
Manual removal works best for low volumes or very short documents. It requires a trained eye for pronouns and gendered language and constant vigilance. Many recruiters find it slows down their workflow significantly.
Method 2: Scripted / regex removal
Scripts using regular expressions (regex) can automatically scan documents for common pronouns and replace them.
Example regex might catch:
he|she|him|her|his|hers|they|them|their|theirs
Pros:
- Fast and scalable
- Easy to integrate into existing ATS or workflow scripts
Cons:
- Only catches exact word matches (case sensitive or insensitive)
- Misses neopronouns (e.g., ze, xe, hir) unless explicitly added
- Can produce false positives, e.g., "he" inside "the"
- No context awareness. Can remove pronouns that are part of quoted text or company names
Regex is a solid baseline but requires ongoing updates and manual review to catch errors and omissions.
Method 3: AI / ML detection
AI and machine learning models can detect pronouns in context, including less common or newly coined ones.
Pros:
- Contextual understanding reduces false positives
- Can identify gendered language beyond simple pronouns
- Adaptable to new pronouns and languages
Cons:
- Requires training data and ongoing tuning
- Can be a black box. Errors may be hard to explain or fix
- May struggle with international names or mixed-language documents
- More resource-intensive to implement
AI is promising for agencies with high volumes and complex documents but needs careful oversight. It's rarely perfect out of the box and can't yet fully replace human review.
Method 4: Distill (one-click pronoun removal)
Distill automates pronoun detection and removal in one step. It integrates with your workflow to strip all gendered pronouns—including common, neopronouns, and variants—from CVs and related documents before submission.
What Distill does:
- Removes he, she, they, ze, xe, hir, and others
- Handles case, plural forms, and possessives
- Preserves document formatting and readability
- Works across UK, Australian, and European recruitment standards
Tradeoffs:
- Automated removal can sometimes affect sentence flow
- Not a substitute for a full blind hiring policy but a practical step in the process
Distill saves hours of manual work and reduces risk of bias slipping through pronouns in blind hiring workflows.
Edge cases and pitfalls
- International and mixed-language documents: Pronouns vary widely and may not be caught by English-only scripts or AI. Some languages gender nouns or verbs rather than use pronouns.
- Neopronouns: New pronouns appear regularly. Tools need updating or training to keep pace.
- Context matters: Pronouns in quotes, testimonials, or third-party references may not indicate candidate gender. Auto-removal can distort meaning if not checked.
- False positives: Regex scripts can remove "he" inside other words (e.g., "the", "here") without context.
- Candidate preferences: Some candidates specify pronouns deliberately. Removing these may conflict with inclusivity goals if the workflow doesn't clarify intent.
Always review automated removals regularly and combine with broader blind hiring policies.
FAQ
Q: Can pronouns be completely removed without affecting readability? A: Usually yes, but some sentences may sound awkward after removal. Some agencies replace pronouns with neutral placeholders or rephrase sentences manually.
Q: Are gender-neutral pronouns like "they" always removed? A: Yes, to avoid any gender signal. This may remove non-binary pronouns candidates use intentionally. Balance blind hiring goals with inclusivity policies.
Q: How often should scripts or AI models be updated? A: Regularly. New pronouns emerge, and language evolves. Quarterly reviews are a good starting point.
Q: Does GDPR affect pronoun removal? A: Pronouns are personal data under GDPR's broad definition. Removing them matches minimising personal data in blind hiring but doesn't replace consent or data subject rights.
Q: Can Distill integrate with ATS systems? A: Distill formats and cleans CVs before submission but doesn't currently integrate natively with ATS platforms. It works alongside your existing tools.
If you're sending 20+ CVs a week through blind hiring workflows, automating pronoun removal saves time and reduces bias risk. Try Distill free to strip pronouns and other gender signals in one click.