How AI Recruiting is Destroying the Job Market: A Recruiter's Perspective

Here's an uncomfortable truth: AI-written job descriptions are attracting AI-written résumés, that are being screened by AI-driven Applicant Tracking Systems (ATS), that are selecting candidates for video interviews that will use AI to evaluate their facial expressions, body language, and vocal inflection. The system is broken.

AI has become a popular tool in hiring processes across many organizations. It promises faster screening, reduced bias, and more efficient recruitment. Yet, despite these promises, it has complicated the hiring process, bypassed qualified applicants and may be preventing organizations from finding the best available talent.

This isn't an indictment of the technology but rather a way of starting a long overdue conversation few people seem willing to have. There is growing evidence that qualified candidates are being reduced to mere algorithms and data points on a dashboard, while recruiters play a secondary role in a game of cat and mouse to catch those who dare cheat the AI system. This trend is neither beneficial nor sustainable for organizations looking to remain competitive in attracting top talent.

Candidates are increasingly being reduced to a series of algorithms and datapoints in the hiring process.

How AI Shapes Hiring Decisions

AI systems analyze resumes, cover letters, and even video interviews to rank candidates. They use algorithms trained on historical hiring data to predict which applicants fit best. This approach can save time by filtering out unqualified candidates quickly.

However, it depends on the data it learns from. If past hiring decisions favored certain profiles or overlooked others, the system will replicate those biases. For example, if a company historically hired mostly candidates from a specific university or background, the AI might undervalue applicants outside that group.

AI vs. Recruiter

Over the course of my 20+ years in HR, I've conducted somewhere between five to six thousand interviews. I often think about how each experience helped me refine my skills in different ways.  I mastered the ability to quickly establish rapport with total strangers and transform that connection into trust by the end of a thirty-minute phone conversation. I learned how to integrate mundane interview questions into casual conversation and how to tell the difference between an applicant who was unprepared, nervous, or being intentionally deceptive. With each new interaction I learned empathy, compassion, critical listening skills, and more about myself than I ever expected.

I perfected my craft through decades of service alongside some of the finest talent acquisition and organizational development professionals in the nation. My skill set was forged through trial and error. I remember every failed search, rejected offer, and candid discussion with a colleague, mentor, client, or supervisor. Regardless of how the search ended, my goal was always the same—get better. This is where AI falls short.

The Difference Between Information, Knowledge and Experience

The primary issue is that most AI are poor at reasoning, often reaching conclusions with little regard for diplomacy, accuracy, or collateral damage. It will often resort to just telling the user what it thinks they want to hear because it doesn't understand that information alone does not equate to knowledge, and neither one is particularly useful without practical application.

• Information is recognizing candidates whose resumes show brief tenures at each job.

• Knowledge is understanding that those same candidates (with the right skill set) may still have value.

• Experience is the ability to extend the benefit of the doubt to a candidate who may have been the spouse of a service-member frequently required to move, which could account for the short durations at each position.

Lack of Human Judgment and Context

AI tools simply cannot fully understand context or nuances in candidate profiles. They might reject applicants who took career breaks for valid reasons or who have non-traditional experiences or hobbies.

Human recruiters can ask follow-up questions, interpret tone, and assess cultural fit. AI lacks this flexibility and may make decisions based solely on data points and trends, which can be misleading.

One study in particular, Evaluating the Promise and Pitfalls of Using LLMs in Hiring Decisions, found that Large Language Models (LLMs) selected resumes that were AI-generated 95% of the time over the very same resumes that were designed without AI assistance. The system clearly demonstrated a preference towards résumés that utilized its own language, format and font style.

This isn't progress; it's a high-tech return to the biased hiring practices of the past, where top talent could be overlooked simply because their handshake wasn't deemed firm enough or because they avoided direct eye contact with the hiring manager.

Overlooking Soft Skills and Potential

AI excels at evaluating hard skills and keywords but struggles with soft skills like communication, creativity, and adaptability. These qualities often determine long-term success but are difficult to quantify.

For instance, a candidate with an unconventional career path might bring fresh perspectives and problem-solving abilities. The system might rank such a candidate lower because their resume does not match typical patterns. This means organizations risk missing out on diverse talents who could thrive in dynamic roles.

Practical Steps to Improve AI Hiring

Organizations that want to benefit from AI while avoiding its pitfalls can take several steps:

• Regularly audit AI systems to detect and correct biases.

• Combine AI screening with human review to add judgment and context.

• Train AI on diverse data sets that represent a wide range of candidates.

• Use AI to support, not replace, recruiters ensuring final decisions consider human insights.

• Focus on soft skills by incorporating assessments beyond resumes and keywords.

• Deemphasize Time-to-Fill as the defining metric of a successful hire.

Final Thoughts on AI in Hiring

AI can still be a useful tool in recruitment but relying on it exclusively limits an organization’s ability to find the best talent. Bias in data, challenges in evaluating soft skills, and a lack of context can lead companies to miss out on valuable candidates and potentially expose them to future litigation.

As a recruiter, there will always be tasks that AI can perform faster or more accurately than I can, but my strength lies in my ability to identify with my candidates in ways that it can only hope to replicate someday. It cannot comprehend the need for a career break to care for a sick family member or the desire for a mid-career change. It will not understand the pain of losing a job, emotional burnout, declining health or how these events impact candidates' career decisions. It will never fully grasp the human experience, nor should we expect it to. However, we need to acknowledge these limitations and create a set of guiding principles that protect the civil rights of our workforce from being infringed upon.

AI is here to stay, but the use of this technology in modern hiring practices comes with increasing responsibility. We are more than algorithms, dashboards or even the sum of our own knowledge. Our worth to these organizations should not be determined by our skill in crafting the ideal AI-friendly resume or how our body language is evaluated in a video interview. Our value is rooted in our ability to offer creativity, resilience, determination and innovation, which stem uniquely from our human experiences.