AI is helping candidates optimize resumes while employers use technology to screen growing applicant pools. But are hiring systems finding the strongest people to interview or the resumes easiest to recognize?
By Sherri Claydon | WBN News | October 20, 2026
Editor: Karalee Greer Subscription to WBN and being a Contributor is Free
AI is helping candidates optimize resumes while employers use technology to screen growing applicant pools. The question is whether that process is getting better at finding the strongest people to interview, or the resumes that are easiest for the system to recognize.
Employers have a hiring problem that is easy to understand.
BambooHR data shows the average number of applications per job posting nearly doubled from about 46 in 2021 to 95 in 2025. Employers cannot interview 95 people for every role. They need a way to narrow the field.
Candidates face a problem of their own. They are competing for attention in those crowded applicant pools, and AI can help them tailor a resume to the language and requirements in a job description.
Both uses of technology make sense.
But when AI helps candidates optimize the documents going in, while technology helps employers decide which documents rise to the top, there is a question worth examining:
Are we improving our ability to identify the strongest candidates to interview, or our ability to identify the resumes that best match the job description?
The increase in applications is substantial.
BambooHR analyzed more than 72 million applications, one million job postings and about 6.5 million completed hires across its global platform. Applications per posting rose from about 46 in 2021 to 95 in 2025. Over the same period, the hiring rate fell from 4.5% to 2.8%, while completed hires dropped from 1.34 million in 2022 to 1.05 million in 2025.
Those numbers do not mean more applications caused fewer hires. BambooHR also found that employers were filling a larger share of positions internally and becoming more selective about external hiring.
What the numbers do show is that more candidates entering the funnel has not made hiring easier.
Indeed research on higher-salary hiring in Canada and the U.S. found that 56% of employers who said hiring was difficult identified too many candidates as part of the problem. Thirty-nine per cent were receiving more candidates per opening than the year before. Employers pointed to the economy and better algorithmic matching as major reasons for the increase. Indeed also noted that AI tools are making it possible for job seekers to apply to more roles faster.
AI may be contributing to application volume, but the available evidence does not establish that it caused the sharp increase. Economic conditions, fewer available roles, easier online applications and improved job matching can all play a part.
What is clear is that employers have more information to sort through, while candidates have more tools to compete for attention within it.
Recent research shows the pressure building on both sides of the hiring process:
The issue is not simply that candidates are using AI.
The bigger question is whether the resume still works as well as a tool for deciding who should get an interview when the document itself is becoming easier to optimize.
Candidates are routinely advised to tailor their resumes to each opportunity, use terminology from the job description and make relevant experience easy to find.
AI makes that much easier.
It can review a job description, identify important requirements and help a candidate bring related experience forward. Used well, that can be valuable. A strong candidate should not be overlooked simply because they used different terminology or struggled to explain transferable experience.
Employers, meanwhile, have to reduce a large applicant pool to a workable number.
Applicant tracking systems have been doing some version of this for years. Robert Half's Canadian guidance says ATS software can sort and rank applications according to how well they match a profile, often using specified keywords. Its advice to candidates is equally direct: using the same language as the job description can increase the chance of moving forward.
Newer AI-based systems can go beyond simple keyword matching and consider more context. But the basic purpose of screening remains the same.
The employer is not yet deciding who gets the job.
They are deciding who gets the interview.
That is where human judgment can go deeper. A hiring manager can ask questions, understand the context behind someone's experience, explore accomplishments and begin to assess how that person actually thinks.
But first, the candidate has to make it into that smaller group.
This is where the distinction becomes important.
AI can help a candidate make their experience easier for a screening system to recognize. It can identify terminology in the job description, find related experience in the candidate's background and make those connections much more explicit.
The employer's technology is then evaluating that document to help decide who moves forward.
That does not necessarily mean the most optimized resume belongs to the strongest candidate.
Robert Half Canada has described what it calls a "resume illusion," where job seekers use generative AI to create keyword-focused resumes that perform well in screening but the apparent depth of experience does not always hold up in the interview. The company advises employers not to rely solely on resume-screening tools to identify the best candidate.
Research into automated screening raises another concern. A 2026 ACL study tested self-promotional instructions designed to influence LLM-based resume screening without adding new qualifications. In some experimental conditions, the wording improved rankings and occasionally allowed a lower-quality candidate to outrank a higher-quality one. The study examined deliberate manipulation, so it should not be treated as evidence that normal AI-assisted resumes routinely produce the same result.
But both findings point toward the same question.
How well does the strength of the document reflect the strength of the person behind it?
A candidate with deep leadership experience, for example, may describe building teams, developing managers, leading through change and solving difficult business problems. Another candidate may have less depth but use AI to make every example of leadership highly visible and closely aligned to the language in the posting.
A strong screening system may recognize the substance in both.
But the growing ability to optimize a resume makes it worth asking whether the candidates who are easiest for a system to recognize are always the candidates an experienced hiring manager would most want to meet.
We do not yet have enough evidence to say they are not.
What we do know is that employers are increasingly questioning how much the resume tells them.
A resume has never proven that someone can do a job.
Its first job is much simpler: give the employer enough evidence to decide that this person is worth talking to.
That becomes more complicated when resumes themselves are increasingly optimized.
Robert Half's Canadian research found that 64% of hiring managers said the increased volume and authenticity concerns surrounding AI-generated resumes were creating challenges. Employers reported spending more time reviewing applications and adding steps to validate candidates.
Express Employment Professionals and The Harris Poll found a similar confidence problem. Eighty-two per cent of Canadian hiring managers said candidates' resumes at least sometimes fail to match their real-world skills. At the same time, only 22% of job seekers acknowledged listing skills they did not have.
Those findings should not be read as evidence that most candidates are being dishonest.
They show something broader: employers are becoming less confident that the document alone tells them what they need to know.
Criteria and Lighthouse Research & Advisory found only 33% of employers were very confident resumes reflected candidates' true skills. Ninety-eight per cent said they trusted assessments, structured interviews and work samples more than resumes.
The challenge is that those stronger forms of evaluation usually happen after the initial screening.
Candidates who do not make the first cut never reach the stage where an employer can ask better questions, understand the context of their experience or see what they can actually do.
This is the question I keep coming back to in recruitment.
Employers tell me they are struggling to find the calibre of candidates they want. At the same time, experienced professionals are struggling to get interviews.
That observation does not establish that AI screening connects the two. There are many forces shaping today's labour market.
But the research gives us reason to look more closely at what happens between application and interview.
BambooHR says rising candidate volume can become friction when screening signals are noisy. Its 2026 analysis describes a possible "matching freeze," where candidates are entering the funnel but the process itself prevents them from converting into hires.
Canadian job seekers are noticing the screening stage too. An August Express Employment Professionals and Harris Poll survey found 35% said navigating an automated hiring process instead of feeling heard by people was a challenge in their job search. The same study found 32% of Canadian hiring managers had positions they could not fill, while 17% of open positions, on average, were eventually closed without anyone being hired.
None of that proves strong candidates are routinely being screened out by AI.
It does suggest the part of hiring that happens before the interview deserves more attention.
There is also something a resume can lose when it becomes too focused on matching a posting.
For an experienced professional, a career often makes more sense as a whole.
There can be a reason one role led to another. An apparently lateral move may have expanded someone's scope. Someone may have inherited a difficult situation, built a team, developed people or solved a problem that does not fit neatly into a keyword.
Sometimes the sequence is part of what shows the depth of someone's experience.
Generative AI can be useful here. It can improve clarity, help translate transferable experience and identify accomplishments that deserve more attention.
But there is a difference between making someone's experience easier to understand and reshaping it around what a screening system is most likely to recognize.
Research presented at LREC 2026 offers some evidence that AI can also standardize presentation. Researchers studied 420 authentic, AI-enhanced and fully AI-generated resumes across five IT job descriptions and found systematic stylistic differences among the three groups.
That does not mean AI-generated resumes all sound the same.
It does reinforce the importance of separating how effectively experience is presented from the substance of the experience itself.
This is not uniquely a Vancouver issue, but British Columbia is a useful place to examine it.
Statistics Canada reported in June that 25% of B.C. workers had used generative AI at work during the previous year, the highest rate among the provinces and regions reported. Ontario was at 23.6%, Quebec at 21.2% and Alberta at 21.1%. B.C. remained ahead of Ontario even after Statistics Canada adjusted for worker and workplace characteristics.
Greater Vancouver also sits within a large professional labour market.
WorkBC's Mainland/Southwest region, which includes Greater Vancouver, the Fraser Valley, Sunshine Coast and communities farther north, had about 224,700 people working in professional, scientific and technical services in 2024. Almost three-quarters of B.C. employment in that industry was concentrated in the region.
As of August 2026, the Mainland/Southwest unemployment rate was 6.8%.
This is a large professional workforce in a province where workplace use of generative AI is already high.
How employers and candidates use that technology to find one another matters.
AI does not necessarily make the resume obsolete.
It may force us to be clearer about what the resume is for.
If the question is, "Is there enough evidence here to justify a conversation?" the resume can still be extremely useful.
The harder question is how much confidence employers should place in an increasingly optimizable document when deciding who gets that conversation.
The answer is not necessarily more hiring stages. Canadian employers are already reporting longer hiring timelines. Twenty-eight per cent told Express Employment Professionals and The Harris Poll that hiring takes longer than it did two years earlier, while 30% said the process takes at least four weeks.
It may instead mean becoming more deliberate about what each stage is expected to tell us.
Screening can help make a large applicant pool manageable.
The interview can explore context, judgment and the substance behind the resume.
Where appropriate, work samples or assessments can provide another view of capability.
And throughout the process, the technology should support the hiring decision rather than become a substitute for understanding the people being considered.
Candidates are not going to stop using AI to improve their applications. Employers are not going to stop using technology to manage large applicant pools.
Nor should either side necessarily do so.
The more useful question is whether the hiring process is optimizing for the thing employers actually need.
A resume that matches a job description extremely well may deserve attention. But matching the language of a requirement and demonstrating the capability behind it are not quite the same thing.
For employers, the first screening stage does not need to determine who should get the job.
It needs to help identify the right people to talk to next.
That distinction matters because once a candidate disappears from the process, all the human judgment that comes later never gets a chance to happen.
As AI gets better at helping candidates present themselves and employers sort through them, the challenge may be making sure we do not become so good at matching documents that we lose sight of why we were matching them in the first place.
The goal was never to find the best resume. It was to find the person behind it.
By Sherri Claydon | WBN News - Vancouver LinkedIn: https://www.linkedin.com/in/sherriclaydon
Editor: Karalee Greer Subscription to WBN and being a Contributor is Free
Tags: #Sherri Claydon #WBN News #Artificial Intelligence #Hiring #Future of Work #Recruitment
Source Information
