Candidates have AI.
Employers have AI.
And now both sides are adapting to what the other side is doing.
A candidate can use AI to improve a resume, prepare for an interview, research a company or organize a job application.
An employer can use AI to screen resumes, identify skills, rank candidates, automate communication and increasingly handle parts of the recruiting process.
That creates an interesting cycle.
Candidates use AI because hiring systems are becoming more automated.
Employers use more AI because applications are becoming easier to produce.
Candidates then adapt to those systems.
Employers respond again.
And suddenly hiring starts to look less like a straightforward process and more like a technological arms race.
The problem is that neither side necessarily wins.
The Application Has Changed
A resume used to require time.
A cover letter required even more.
If someone wanted to apply for ten jobs, they had to spend a significant amount of time preparing ten applications.
AI has changed that equation.
Candidates can now use software to identify important terms in a job description, reorganize their experience, improve wording and prepare different versions of an application much faster.
That can be useful.
A candidate who struggles with writing can communicate their experience more clearly.
Someone changing industries can identify transferable skills.
Someone applying internationally can make their experience easier for an employer to understand.
The problem appears when the application becomes easier to produce than the underlying qualifications.
Suddenly, employers may receive a large number of polished applications without knowing how much information in those applications actually helps predict performance.
That changes what a resume is worth.
Employers Have a Volume Problem
This is one reason employers are turning to AI.
Greenhouse's 2026 hiring benchmarks analyzed more than 640 million applications from more than 6,000 companies and found that applications had surged while recruiting teams were handling substantially more applications.
When the number of applications increases, somebody has to process them.
That can mean more recruiters.
Or it can mean more technology.
AI can scan information much faster than a person can.
It can identify skills.
It can organize applicants.
It can flag potential matches.
It can help recruiters decide where to spend their attention.
From an employer's perspective, that makes sense.
But it creates a new incentive for candidates.
If they know an automated system is screening applications, they have a reason to optimize their applications for that system.
And the cycle begins.
Candidates Learn to Optimize for the Machine
Candidates have always adapted to hiring systems.
When employers started using applicant tracking systems, candidates learned to pay more attention to keywords.
When companies started asking for specific certifications, candidates highlighted those certifications.
When employers started emphasizing skills, candidates began putting more evidence of those skills into their resumes.
AI simply makes this adaptation faster.
A candidate can ask an AI system:
What skills does this job description emphasize?
Which parts of my experience are most relevant?
How should I explain this project?
What questions might come up in an interview?
Again, none of that is automatically dishonest.
The line becomes much more important when AI is used to invent experience, disguise a lack of knowledge or deliberately manipulate an automated screening system.
At that point, the employer has a new problem.
The system is no longer simply evaluating candidates.
It is being played by them.
Employers Start Looking for Proof
Once employers become concerned that AI is influencing applications, they have an obvious response.
Ask candidates to prove themselves in another way.
That could mean:
- a live interview
- a practical assignment
- a work sample
- a skills test
- additional questions
- identity verification
- a live demonstration
- a request to explain how a piece of work was created
Some of these are perfectly reasonable.
In fact, there is a strong case for assessing what someone can actually do rather than relying entirely on a resume.
NACE's 2026 research found that 70% of employers surveyed use skills-based hiring, up from 65% the previous year. Employers most commonly use this approach during interviews and screening.
That is an important shift.
It means employers are increasingly interested in evidence of ability rather than simply credentials.
But there is a difference between testing real skills and trying to catch candidates using AI.
The Problem With an Endless Verification Process
Imagine what happens if employers respond to AI-assisted applications by adding more and more checks.
The candidate submits a resume.
Then completes a questionnaire.
Then records a video.
Then completes an assessment.
Then joins a live interview.
Then explains the assessment.
Then completes another assignment.
Then verifies their identity.
The employer may feel more confident.
The candidate may feel exhausted.
And if several companies are doing the same thing, strong candidates have another option:
They can simply stop applying.
This is where the arms race becomes counterproductive.
The technology is supposed to make hiring more efficient.
Instead, both sides can end up spending more time trying to protect themselves from the other side.
The Real Problem Is Trust
At its core, this isn't really an AI problem.
It is a trust problem.
Employers want to know:
Is this candidate really as capable as this application suggests?
Candidates want to know:
Will I actually be evaluated fairly, or am I just trying to get past an automated filter?
Both questions are reasonable.
The trouble starts when neither side trusts the process.
The employer responds with more screening.
The candidate responds by optimizing more aggressively.
The employer adds another verification step.
The candidate prepares for that step.
Eventually, the hiring process becomes a competition to outsmart the other side.
That is not what hiring is supposed to do.
Employers Should Ask a Better Question
Instead of asking:
“How do we stop candidates from using AI?”
employers could ask:
“How do we design a hiring process that shows us whether someone can actually do the work?”
Those are very different questions.
Someone can use AI to improve a resume and still be an excellent employee.
A marketer can use AI for research and still be a strong strategist.
A writer can use AI for brainstorming and still produce original, thoughtful work.
A developer can use AI to generate code and still understand the system they are building.
AI use does not automatically tell an employer whether someone is qualified.
The quality of the work does.
Give Candidates Real Problems
This is where practical hiring becomes more valuable.
Instead of creating a complicated test designed primarily to prove that someone is not using AI, employers can give candidates a realistic problem connected to the actual job.
A marketing candidate could review a campaign and explain what they would change.
A customer support candidate could respond to a realistic customer situation.
A salesperson could explain how they would approach a difficult prospect.
A writer could improve a short piece of content and explain the changes.
A project manager could describe how they would handle competing deadlines.
The point is not to create another obstacle.
It is to see how the person thinks.
If the candidate uses AI along the way, that may not even be the most important question.
Can they explain the result?
Can they spot mistakes?
Can they defend their decisions?
Can they improve the work when challenged?
Can they apply judgment?
Those are much more useful signals.
Candidates Have a Responsibility, Too
The arms race is not only an employer problem.
Candidates need to think carefully about how they use AI.
There is a reasonable difference between:
“Help me explain what I actually did.”
and
“Invent something I never did.”
There is a difference between preparing for an interview and having AI answer every question for you.
There is a difference between improving your writing and pretending to have skills you do not have.
AI can make a job search easier.
It cannot replace the ability to do the job once you get hired.
Remote workers are also thinking differently about the opportunities available to them. Some may even consider taking on more than one remote role, depending on their contracts, schedules and workload. We looked at what workers should consider before doing that in our guide to working two remote jobs at the same time.
That distinction matters because the hiring process eventually ends.
The work begins.
Skills May Matter More as AI Gets Better
There is another interesting consequence of this shift.
As AI makes it easier to produce polished applications, employers may place more weight on evidence of actual skills.
That is already happening.
NACE found that 70% of employers in its 2026 survey reported using skills-based hiring. The organization also found that employers emphasize problem-solving, communication and teamwork alongside job-specific skills.
This creates an interesting future for candidates.
The strongest application may not be the one that sounds the most impressive.
It may be the one that gives the employer the clearest evidence that the person can actually do the work.
That could mean portfolios.
Projects.
Work samples.
Specific examples.
Practical interviews.
References.
And conversations where candidates can explain how they approach problems.
In other words, hiring may become less about producing the perfect application and more about demonstrating something real.
AI Skills Are Becoming Part of the Job, Too
There is another side to this story.
Employers are not only trying to figure out whether candidates use AI.
Increasingly, they want candidates who know how to use it well.
NACE's 2026 research found that demand for AI skills in entry-level jobs had risen sharply, with employers increasingly looking for candidates who can use AI tools appropriately, evaluate AI output and apply those tools to their work.
That creates an important distinction.
The future hiring question may not be:
“Did you use AI?”
It may be:
“What did you use AI for, and did you know when its answer was wrong?”
That is a much more useful question.
We Don't Need to Win the AI Arms Race
There is a temptation to solve every new hiring problem with another piece of technology.
AI helps candidates apply faster.
So employers use AI to screen faster.
Candidates optimize for screening.
Employers develop better detection.
Candidates adapt again.
Employers add verification.
And the cycle continues.
But hiring does not need to become a technological competition.
Employers do not need to catch every person who uses AI.
Candidates do not need to prove that every word of their application was typed without assistance.
Both sides need something simpler.
A reliable way to understand whether the person and the job are actually a good match.
That requires technology.
But it also requires judgment.
The Best Hiring Process May Be the One That Needs Less Proof
A strong hiring process should make it difficult to fake the things that actually matter.
If communication is important, have a real conversation.
If writing is important, review real writing.
If technical ability is important, look at real technical work.
If problem-solving is important, give the candidate a problem worth solving.
If experience matters, ask about what they actually did.
The goal is not to eliminate AI from hiring.
That is probably unrealistic.
The goal is to make AI less important to the final decision.
Because once employers can see what a candidate actually knows and can do, it matters much less whether AI helped them get the resume into shape.
Hiring Is Still About People
The AI hiring arms race is interesting because both sides are trying to solve the same basic problem.
Employers want better information about candidates.
Candidates want a fairer way to demonstrate what they can do.
AI can help both sides.
It can also make the process more complicated if everyone starts using it to protect themselves from everyone else.
The answer is not to stop using technology.
It is to use it more deliberately.
Let AI handle repetitive work.
Let candidates use tools that help them communicate clearly.
But keep the parts of hiring that require judgment, context and human interaction firmly in the process.
The winners of the AI hiring race may not be the companies with the most sophisticated screening systems or the candidates with the best prompts. They may be the people on both sides who learn how to use AI without letting it replace good judgment.
Looking for Your Next Remote Job?
AI may be changing how employers hire, but real skills, experience and judgment still matter.
If you're looking for a remote opportunity, explore current openings on Online.jobs and find roles that match your skills and experience.
Sources
- Greenhouse, The Hire Standard: 2026 Hiring Benchmarks — analysis of more than 640 million applications across 6,000+ companies. Read the Greenhouse hiring benchmarks
- NACE, Employer Use of Skills-Based Hiring Practices Grows — 2026 employer research on skills-based hiring. Read the NACE research
- NACE, Ready or Reluctant? Employer Expectations for AI Skills Meet Student Skepticism — 2026 research on AI skills in early-career hiring. Read the NACE findings
- SHRM, The State of AI in HR 2026 — research based on 1,908 HR professionals and AI adoption in HR. Read the SHRM report