What If AI Could Help You Screen People In Instead Of Out?
To our Talent Community,
Over the last six months, we've been writing about AI's transformation of recruiting tools, processes, and strategy. As job markets have rebounded throughout the last year, one bet we see companies consistently making is applying AI at the top of the funnel.
Teams are turning to AI to reduce sourcing efforts, cut recruiting time, and speed up the process of finding top inbound talent, but that’s not the only significant benefit from applying AI at the top of the funnel.
According to Applied's research, using “traditional” screening methods, 60% of people hired would have been missed via traditional CV screening. Meanwhile, hiring statistics show that 75% to 88% of job applicants are considered unfit for positions, and only 2% of applicants make it past the resume screening stage.
This raises the question: is the problem the candidate pool, or the screening process? The evolution of AI presents a unique opportunity—instead of filtering out candidates, modern AI can help find hidden talent pools that traditional methods miss. We can think of this as the “screen in” revolution.
The Problem with Traditional Screening
Before talking about how AI can screen people IN, we need to understand why previous methods and tools screen so many qualified people OUT.
Boolean and keyword searches miss qualified talent. These searches can't identify candidates with transferable skills. They overlook passive candidates who don't have detailed experience descriptions and they can create false positives due to poor signal quality from LinkedIn skill tags. They can often miss skills that experts could infer without specific keywords.
Human bias eliminates talent before evaluation. Research shows that 80% of candidates are filtered out based solely on current company or university. This limits the talent pool to familiar brands and credentials. Especially when pressed to fill roles quickly, recruiters can often focus narrowly on current job titles and their associated skill sets, missing the broader transferable skills that predict success across roles and the potential for growth into future positions.
Limited recruiter knowledge constrains sourcing. Recruiters spend about one-third of their work week sourcing candidates, yet 72.8% of recruiters struggle to find relevant candidates. This creates blind spots in sourcing strategies. The result is repeatedly sourcing from the same familiar talent pools while missing high-potential candidates from industries, companies, or educational backgrounds that aren't on the recruiter's radar.
Old AI vs. New AI in Recruiting
Pre-2023 Bias Amplification to Post-ChatGPT Objective Evaluation
Many companies who remain slow to adopt AI or apply it only peripherally cite fear that AI will introduce or perpetuate bias. This concern is shared by both candidates and company leaders—34% of people see AI hiring tools as more prone to bias than humans alone and we frequently see HR executives taking a “wait-and-see attitude” due to fear of AI's unknown repercussions.
While it's critical to take precautions and establish responsible boundaries between AI and human input, recent advancements in AI are now helping address long-standing concerns about bias in the recruiting process. (Warning: we’re going to get a little nerdy here!)
Prior to 2023, AI tools leaned heavily on machine learning, which relied on historical data which could in fact reinforce existing biases. As an example, Amazon's infamous AI recruiting tool, scrapped in 2018, systematically discriminated against women. The system was trained on 10 years of resumes from male-dominated tech roles. It penalized resumes with words like "women's" and downgraded candidates from all-women's colleges.
More recently, Workday faces a collective action lawsuit alleging its AI-powered screening tools discriminate against applicants over 40. The case, which gained class-action status in 2025, claims that Workday's algorithm "disproportionately disqualifies individuals over the age of forty from securing gainful employment," with some applicants receiving rejection emails within minutes or hours of applying.
Many old AI systems used static approaches and couldn't adapt to changing hiring needs or role requirements.
The evolution from traditional machine learning to Large Language Models (LLMs) represents a fundamental shift in how AI processes information. This evolution has important implications for mitigating bias in recruiting, especially when these models are purposefully tuned for recruiting scenarios and equipped with the contextual understanding needed to think and reason about candidate evaluation in a more nuanced, thoughtful way.
The Technical Evolution: Earlier AI systems treated resumes as a collection of individual words, scanning for keywords without truly understanding context. This often led to missed connections across a candidate’s experience. The 2017 breakthrough in "transformer" technology marked a shift—enabling AI to better grasp the relationships between words and interpret meaning across sentences, more like how a human would read and understand a resume.
How This Reduces Bias: Research on LLMs shows they are "robust across race and gender" compared to traditional machine learning models. The self-attention mechanism allows modern AI to focus on job-relevant skills and experiences rather than getting stuck on superficial patterns that earlier models would latch onto.
Modern Capabilities: One emerging application of today’s LLMs—popularized by platforms like Endorsed—is the use of customizable evaluation criteria, enabling companies to assess candidates based on core skills rather than credentials and to tailor evaluations to specific role requirements in real time. These tools can analyze resumes and LinkedIn profiles for relevant keywords, skills, and experiences, supporting a shift toward more skills-based hiring. Importantly, this is just one of many possible applications of LLMs in recruiting. As the technology continues to advance, entirely new approaches are on the horizon — offering greater automation, intelligence, and potential for reducing bias in candidate evaluation.
The Key Difference: Where traditional machine learning models learned rigid patterns from historical data, transformer-based LLMs can understand context and nuance. This allows them to evaluate candidates based on actual qualifications rather than proxy signals that may correlate with protected characteristics.
How AI Screens People “In”
Applying LLM tools to top-of-funnel recruiting doesn't just potentially reduce bias—it can fundamentally expand how recruiters and hiring managers think about what makes a great candidate and unlock entirely new ways to identify candidates who could excel in the role.
AI identifies transferable skills across industries and discovers high-potential candidates from lesser-known organizations. As an example, AI can recognize that a data analyst at a nonprofit has statistical analysis and reporting skills directly applicable to business intelligence roles in retail or finance, while traditional screening focuses only on sector-specific experience.
AI moves beyond keywords. Endorsed uses "full market sourcing,” a new approach where recruiters start with thousands of candidates and let AI review each one individually, then calibrate based on what they like, rather than traditional keyword filtering. This LLM-powered method finds 2-3 times more relevant candidates by understanding implicit skills, researching companies, and identifying quality matches that keyword searches miss.
Expands sourcing beyond recruiter knowledge limits. AI can search databases and platforms that recruiters might not know about. It can identify relevant companies and talent pools automatically. For example, AI might discover talented developers through their contributions on Stack Overflow, GitLab, or Kaggle competitions, or identify researchers through their publications—platforms that many recruiters don't actively monitor but contain rich talent data.
Helps define role requirements. A key challenge with screening-in is knowing what you need ahead of time. Well-defined roles and structured interviewing enable better screening. AI can help educate hiring teams about new roles and required skills. In the past, this required having conversations with people or burning through candidates to learn. AI can fill this educator role and set up searches to be more intentionally inclusive.
Shows its work and creates learning loops. Modern AI can explain exactly why it selected each candidate based on specific requirements, unlike traditional screening where decisions often lack clear rationale. This transparency saves recruiters time documenting reasoning across hundreds of profiles while building trust through clear explanations. It also creates a bi-directional learning opportunity where recruiters can review AI decisions, provide feedback, and help the system improve its understanding of what predicts success in specific roles.
Real-World Impact and Data
Talent teams implementing modern AI screening tools are seeing dramatic improvements in both efficiency and diversity outcomes:
Candidate pool expansion. LinkedIn data shows that focusing on skills can increase talent pools by 10x. Tools like Gem and Findem's attribute-based search can find diverse candidates that traditional keyword searches miss, and platforms like Transformify ensure no candidate is automatically screened out, with AI providing insights to support human decision-making rather than replacing it.
Time and cost savings. Companies using AI in recruitment see a 50% reduction in time-to-hire. AI-powered hiring tools can reduce recruitment costs by up to 30%. For example, Endorsed allows recruiters to give plain English instructions like "find candidates with 3+ years of security engineering experience" and automatically screens thousands of resumes and public profiles to surface the best matches with explanations. Gem's AI App Review instantly ranks hundreds or thousands of applications by analyzing job descriptions and providing detailed match scores, allowing recruiters to process in minutes what previously took hours or days of manual review.
Quality improvements. Companies using AI recruitment tools report 82% better quality hires. AI-picked candidates are 14% more likely to pass interviews and 18% more likely to accept job offers. Eightfold's platform uses deep learning to match candidates based on skills and potential rather than just credentials, and resurfaces past candidates who were screened out previously but may be a fit for current roles.
Diversity gains. AI-powered recruitment tools show a 20% increase in hiring of underrepresented candidates. Tools like Greenhouse and Eightfold offer anonymous screening to ensure human decision-makers choose candidates based on qualifications, not background. Companies implementing these approaches report not only more diverse hires but also improved performance outcomes, as they're selecting based on actual capability rather than demographic assumptions.
Implementation Best Practices
Successfully implementing AI for top of funnel screening isn't just about choosing the right technology—it's about creating a comprehensive strategy. Organizations must approach AI recruiting thoughtfully:
Choose AI tools that prioritize inclusive screening. Look for vendors that can demonstrate bias testing and inclusive design principles. Tools like Endorsed train their algorithms using plain English instructions rather than relying on black-box models, making it easier to understand why a decision was made. The most effective tools also offer explainable AI outputs, allowing users to verify and interrogate results using the “trust but verify” approach. Features like built-in diversity indicators and candidate masking can further help reduce bias and support more equitable hiring practices.
Screen in your own employees. Before posting externally, use AI to screen your existing workforce for role matches based on skills, experience, performance, and potential. AI can identify employees who might excel in different departments, suggest internal candidates for promotions, and help create talent development pipelines that reduce external hiring costs while boosting employee engagement and retention.
Invest in training and change management. 46.2% of companies face technical difficulties integrating AI. Prepare your team for the transition by addressing concerns about job displacement, showing how AI enhances rather than replaces human judgment, and provide clear guidelines on when to rely on AI versus human decision-making. Regular training sessions and feedback loops will help your team adapt to new workflows more effectively.
Pair AI screening with structured interviews for optimal results. AI tools perform best when integrated with structured, consistent interview processes. At Growth by Design Talent, we regularly help clients develop structured interviewing frameworks, and these projects provide perfect opportunities to thoughtfully integrate AI tooling and ensure the entire hiring process works efficiently from screening through final selection.
Balance automation with human judgment. While 70% of employers plan to use AI without human oversight, 75% of people would accept AI decisions only if humans remain involved. Be sure you’re building clear AI operating principles to dictate when, where, and how you’re defining graduated autonomy between AI and humans. Companies like Hubspot are intentionally creating principles that ensure AI isn’t just creating efficiencies, but opportunities for improved human connection and better decision making.
Regular bias auditing and adjustment. AI recruiting is not a "set it and forget it" solution. New York City's 2023 law requires annual third-party bias audits for AI hiring tools, but TA teams should also perform their own due diligence to ensure there are checks and balances. Start by auditing your training data for diversity and representation across demographic groups. Test how your AI performs across different demographic groups using statistical analysis to detect disparities
The Future of “Screening In”
The AI recruiting market is projected to reach $1.12 billion by 2030, growing at a 6.78% annual rate and 87% of companies now use AI-driven tools in their hiring processes.
At Growth By Design, we've seen more and more clients ask for help in moving from traditional screening methods to AI-powered inclusive, efficient hiring approaches. Much of our work focuses on developing comprehensive strategies that expand talent pools while maintaining quality standards. Recently, this has included creating AI operating principles that define when humans should remain in the loop, building structured interview frameworks that complement AI screening, and enabling teams to identify transferable skills across industries. The companies we support consistently see expanded candidate diversity and reduced time-to-hire while discovering talent they would have previously overlooked. If we can be helpful to your team, please reach out to hello@gbdtalent.com.
AI's promise in recruiting isn't about replacing human connections. It's about creating more opportunities for those connections to happen with the right people. When done right, AI doesn't just screen people in—it opens doors that were previously closed. The goal isn't to automate human judgment away, but to give humans better tools to find great people they would have missed otherwise.
—
Mike, Adam & Jill, & special thanks to Allison Slater Rasch for her contributions to this month’s newsletter
🔍 Client Spotlight: Profound
We’re thrilled to be working with Profound, an AI start-up based in New York City that’s building their first internal recruiting team. We’re hiring for two senior/staff-level recruiters—one to lead GTM recruiting and one to lead Tech Recruiting. These folks will be in-office (5 days/week) in Midtown.
There’s lots to be excited about here! Profound has a tight-knit culture that works hard and supports wins as a team. Their Series A was led by Kleiner Perkins back in March 2025, and they’re growing rapidly. The ideal candidate is someone who has an eye for the top 1% talent, strong recruiting operations hygiene, and is well-networked in the NYC area.
If you (or someone you know) are in the NYC area and interested in learning more, please reach out to applications@gbdtalent.com.
🚀 July Leadership Moves in the Market
💗 Holly Dyche, former Principal Talent Advisor at Growth by Design Talent, joined Chainguard as Recruiting Operations Leader.
Jeremy Galossi, former Leadership Recruiting at Character.AI, joined Stainless as Talent.
Justin T. Chen, former Technical Recruiting at Coinbase, joined Succinct as Head of Talent.
Harrison Hernandez, former Head of Recruiting Operations at Plaid, joined Thyme Care as Senior Manager of Recruiting Operations.
BJ McGuire, former VP Global Talent Acquisition at Procore Technologies, joined UKG as their Vice President, Global Talent Acquisition.
Sam Dore, former Head of Talent at Madrona Venture Labs, joined Amplify Partners as a Member of the Talent Team.
Christine Oliver, former Sr. Director, Tech Recruiting at Meta, joined Atlassian as Head of Recruiting, R&D.
Cody Schrotel, former Head of Talent and People Analytics at Squarespace, joined Anaconda, Inc. as their VP, Talent Acquisition.
Derrick Malone Jr., former Global Talent Partner Consultant at a data & tech cloud startup, joined Agoda as Senior Global Talent Acquisition Partner.
Kim Greenia, former Director, Talent Acquisition at Pure Storage, joined Proofpoint as Director, Talent Acquisition.
Maggie Landers, former Vice President Talent Acquisition at Intercom, joined Harvey as VP, Talent.
Andy Thompson, former ops / people / talent at spawn.co, joined Replit as Sr. Director, Recruiting.
Camille Conrotto, former Recruiting at Niantic, Inc., joined Hedra as Head of Talent Acquisition.
Jerry Sastri, former Talent Acquisition at Airbnb, joined Dataiku as G&A Talent Acquisition Manager.
Tina Wig, former Head of Global Talent Acquisition at Moveworks, joined Microsoft AI as Director of Talent Acquisition - Copilot.
Dana Dillard, former Director, Recruiting and Talent at Collective Health, joined Zoox as Director, Talent Acquisition.
James Herriotts, former Senior Director - Talent Acquisition EMEA at The Trade Desk, joined Oviva as Senior Director, Global Talent Acquisition.
Ben Blundell, former Member of Recruiting Staff at OpenAI, joined Meter’s Recruiting, AI models team.
Grant Rivas, former Manager, Portfolio Advisory & Operations at General Catalyst, joined Antimetal as Head of Operations.
Peter Oh, former Director, Talent at Charlie Health, joined Verse Medical as Head of Talent.
Drake Ong, former Vice President, Talent at HUMAN CAPITAL, joined Anduril Industries as Executive & G&A Talent.
Erin Herrera, former Advisor at People Tech Partners Talent, joined Fanatics Commerce as Head of Global Talent Acquisition.
Sierra Kaslow Hedderich, former Talent Leader at Koza Talent Partners, joined Portola as Head of Talent.
Lauren Babek, former Sr. Director, Global Technical Recruiting + University Programs at DoorDash, joined Instacart as Head of Talent.
Conor Sweeney, former Global Head of Talent & Development, joined Form Health as Vice President, Head of People & Talent.
Did we miss your career change? Let us know: hello@gbdtalent.com
📖 What We’re Reading & Listening To
Interviewer Training for Consistent, Quality Hiring - Ashby Webinar featuring our very own Allison Slater Rasch
Zapier’s Bold Bet: Every New Hire Must Be AI-Transformational, With CPO Brandon Sammut - HR Heretics Podcast
Special Episode: LIVE from the HIGHER Altitude Summit, San Francisco - Hiring On All Cylinders Podcast
State of AI Bias in Talent Acquisition - Warden AI
The Impact of AI on the Future of Work - Laszlo Bock
AI Eats the World: Benedict Evans on What Really Matters Now - The MAD Podcast with Matt Turck
Future of Work with AI Agents - Stanford
How to Assess Job Candidates for AI-fluency - Brainfood Live On Air
Why AI Harm To Jobs and Humanity are Vastly Over-Hyped - Josh Bersin
What Gets Measured, AI Will Automate - Harvard Business Review
AI for Recruiters in 2025: What Gets Automated, What Stays Human - The Principal Recruiter
We analyzed 4 million recruiting emails - Steve Bartel, Gem
HR’s role in AI transformation: A playbook for HR leaders - Jacqui Canney, ServiceNow
Perplexity’s CEO on why the browser is AI’s killer app - The Verge’s Decoder podcast
Guidance for candidates on using AI in interviews and hiring processes from Canva and Anthropic
📚 Playbooks:
Your Guide to Evaluating and Selecting an ATS: A guide with helpful process frameworks, and evaluation templates to help you make an informed decision about selecting the ATS that is the best fit for your company.
Navigating an IPO: A guide to help you navigate your team through the IPO process. It includes tools and tactics to prepare your team, proactively communicate to pending candidates, and evolve your recruiting operations to support the shift in business.
Credit for the preview/thumbnail image: David Foodphototasty, Unsplash