How AI Is Making Today’s Recruiting Impossible (How to handle AI-generated resumes, applicant volume, and perfect resumes)

62% of resumes have AI-generated content, so 78% of companies must now check for AI use.

It’s not an exaggeration to say that today, AI is creating havoc in multiple areas of recruiting. In this article, I will focus on the three most impactful strategic areas of corporate recruiting. These areas include the fact that AI usage has skyrocketed, job applications have surged, and the number of perfect resumes has increased dramatically.

During the remainder of this article, I will go into more detail about why each of these three areas is costly. I will also offer some solutions in each problem area that can help limit the negative recruiting impacts of AI.

Problem I – AI Usage Has Skyrocketed, Which Makes Identifying Genuine Talent Among Your Applicants Difficult

The percentage of applicants who are using AI to create their resumes has increased to over 60%. I project that the usage number will soon reach over 80%. Unfortunately, I have found that most of those involved in recruiting are only slightly concerned about the use of AI-generated resumes until they learn the costs of the many problems associated with this emerging tool. 

Why resumes with AI-generated content are problematic

There are numerous reasons why you should consider discounting the resumes of candidates that contain significant AI content. They include:

  • Using AI may reflect on a candidate’s character flaws – a candidate who would secretly utilize AI on a document that is presumed to be written by the candidate. Certainly shouldn’t be considered as transparent or forthcoming. If an applicant submits their resume when they know that it contains exaggerations, omissions, and even mistruths, that candidate could be judged as too tolerant of half-truths and misrepresentations. Using this AI tool may also mean that the candidate may lack energy or be resistant to taking on difficult tasks. And at many companies, any of these potential character flaws would be disqualifying.
  • AI content frequently exaggerates the candidate’s qualifications – as a general rule, AI-generated resumes expand upon or exaggerate the candidate’s capabilities, experience, and qualifications. And in many cases, their resumes include AI hallucinations (i.e., falsehoods that never happened).
  • The use of AI doesn’t show a deep interest in your company – at many companies, it’s a dealbreaker when a candidate doesn’t show a deep interest in your company. And of course, using a single one-size-fits-all resume to apply to all jobs, and allowing an AI agent to determine where their resumes will be submitted. Clearly revealed that the candidate doesn’t have a special interest in your company. If they did, they would have taken the time to tailor their submitted resume specifically to your company.
  • AI-generated content isn’t authentic – many companies are looking for authenticity in a candidate. And literally no one argues that AI-generated content is authentic. This makes a candidate’s resume misleading because it doesn’t reflect the real, authentic candidate.
  • Many hiring managers won’t even consider those who use AI – some companies immediately eliminate candidates with AI-generated resumes, for the simple reason that many of their hiring managers will refuse to give the candidate any further consideration after they have learned about their use of AI.
  • Potential solutions to the AI content problem – the best ways to identify resumes with AI content include looking for major inconsistencies between the applicant’s LinkedIn profile and their resume. Also, keeping track of common ChatGPT resume phrases that are frequently used in resumes can help spot AI-generated resumes. Finally, utilizing AI detection software like GPTZero while at the same time warning potential applicants that you won’t accept resumes with AI content can also be effective.

Problem II – Job Applications Have Surged, So Sorting Through This Volume Is Difficult

It’s a fact that the volume of job applications has surged a whopping 45% this year. As more job seekers learn about AI agents, the number of spam resumes will surely increase. And this increased volume of applications will create several serious problems for the recruiting function. Including:

Why sorting through a large volume of applicants difficult

Some of the problems that an increased volume of applications can create in recruiting include:

  • More resources will be required – such a large volume of applications will take more time and resources, and may even require a new ATS system.
  • More sorting errors will occur – clearly, this increase in volume will overwhelm your existing sorting process, and that will increase the number of your sorting errors. Obviously, more sorting errors will lead to many more unjustified candidate eliminations. And more bad hires.
  • One-size-fits-all applications will become more common – this increased volume will continue to include a large percentage of AI agent-submitted resumes. Most of these will need to be rejected for not meeting your qualifications because these resumes won’t be personalized to your company or job qualifications.
  • Most applicants won’t meet the “deep interest” standard of your company – at corporations where candidates must show a deep interest in your company before they can be hired. An AI-agent submitted one-size-fits-all resumes won’t ever meet the hiring standards. Most candidates don’t even know where these resumes are being submitted. And outside of the content of their cover letter. And at the resume stage, it is difficult to determine the level of an applicant’s interest in the company.
  • Potential solutions to the resume volume problem – are to speed up resume sorting and to make it more accurate. Involve developing a process for quickly identifying resumes with AI content, along with a process that identifies the resumes submitted by AI agents. Finally, you should have a process for quickly eliminating applicants who didn’t reveal a deep interest and knowledge of your company in their cover letter.

Problem III – The Percentage Of Perfect Resumes Has Skyrocketed, So Finding Genuine Talent Is Difficult

When an applicant chooses to target a specific job and company, they can harness AI to generate a resume that hits all of the keywords and the required qualifications in that company’s job posting. These tailored AI-generated resumes are known as “perfect resumes.” And of course, in the past, they were quite rare and were a great find. However, now with the help of AI, both the number and the percentage of “completely qualified” perfect applicants are growing exponentially.  So sorting through them to separate the genuinely exceptionally qualified candidates from the phony ones has become critical.

Here are the problems that perfect resumes can create

This problem area becomes quite serious when the number of perfect resumes that are received exceeds more than a handful. The problems associated with a volume of perfect resumes include: 

  • Creating legal issues – because eliminating a clearly qualified (on paper) candidate has legal ramifications. So you will need to conduct at least some preliminary research into their real qualifications before you drop one of these phony, perfect candidates.
  • You will miss out on actually qualified candidates – when you give interview slots to these “phony” perfect candidates. You will clearly miss out on the opportunity to learn more about those who actually meet your real qualifications (except perhaps for a single minor flaw). 
  • You will encourage more of this bad behavior – if you routinely allow these phony, perfect candidates into the interview step without being challenged. You will inadvertently encourage future applicants to also use AI to make their resumes perfect.
  • Potential solutions to the perfect resume problem – because in reality, the number of perfect resumes that you should receive on an open job should normally be zero. It’s best to automatically freeze any movement of a perfect applicant up to the interview round. At least until after their LinkedIn profile has been checked to verify their resume qualifications. And/or after at least one verification telephone interview has been conducted.

Isn’t it a bit contradictory to actively recruit those who know how to use AI? While at the same time, proactively screening out those who have used AI in their job search?

Final Thoughts

Over the last year, it has become clear that several of AI’s multiple capabilities are now negatively impacting the ability of the recruiting function to accurately sort through its applicants.

I have also recently written about how AI will soon kill remote job interviews. So it’s important to look into the near future. You can expect AI’s capabilities within the recruiting function to continue its upward trajectory towards dominance in sourcing, resume assessment, sorting, and interviews and interview assessment.

And don’t be surprised when AI begins to make its mark on both improving the reference checking process and candidate skills assessment, because AI is currently a wild ride that will only get more intense during the next year!

Note: Next week on 10/13/25 I will publish a companion article, “How to identify AI generated resumes,” which details the best ways to effectively identify the AI generated resumes that you receive.

Additional notes for the reader

This is the latest article from Dr. Sullivan, who was called “the Michael Jordan of Hiring” by Fast Company.

You can subscribe to Dr. Sullivan’s weekly Talent Management articles here or by following him on LinkedIn.

About Dr John Sullivan

Dr John Sullivan is an internationally known HR thought-leader from the Silicon Valley who specializes in providing bold and high business impact; strategic Talent Management solutions to large corporations.

Check Also

Meeting, Conference, Laptop.

AI’s next casualty… Remote Job Interviews (Why AI cheating is bringing them to an abrupt halt)

A whopping 20% of candidates use AI to improve interview scores, and it may exceed …