How to ensure AI tools aren’t filtering out good IT candidates

AI tools aren’t perfect for every job—especially when it comes to hiring staff.

Recruiters are increasingly turning to automation tools to pull prospects from an ever-growing pile of résumés. While that has sped up the hiring process, it has also introduced a fair degree of uncertainty: a 2026 report from hiring platform Greenhouse of 1,865 US respondents (including 1,200 job-seekers and 665 recruiters and hiring managers) revealed that just 21% of recruiters are “very confident” their systems aren’t rejecting qualified candidates.

There are also persistent fears that AI is leading to biased outcomes in hiring. In another study, researchers from Stanford University, Chapman University, and Northeastern University analyzed 4 million job applications and found that automated assessment produced unequal outcomes across racial groups.

If that wasn’t bad enough, those researchers also found that “people who submit multiple applications to positions screened by the same algorithmic hiring vendor are more likely to be rejected from every position to which they apply than would be true if the companies made decisions statistically independently from one another.” For IT pros hunting for work, in other words, algorithms can work against you if you’re applying to jobs at multiple companies.

We spoke with people officers, CIOs, workplace consultants, and HR leads about the best ways for hiring managers and IT leaders to ensure the best candidates get through the application process.

Work with me here. Before consultancy West Monroe used recruiting platform Gem and LinkedIn to bring in candidates, a group including the company’s Chief People Officer Tanya Moore and the talent-acquisition team ran a six-month parallel test comparing Gem’s AI sourcing bot with West Monroe’s own manual selection methods. After finding that the automation left out candidates preferred by human reviewers, Moore and talent acquisition members worked directly with Gem to tweak the algorithm to match their standards.

Moore said her team likes to work with smaller firms like Gem: “Companies that are kind of on the edge of AI but aren’t so big that they really are still looking to partner with their clients.” (Moore said West Monroe runs similar methodology tests twice a year.)

For all companies trying to use AI to hire the best, that continuous audit of rejections is essential, according to CEO and founder of workforce and AI governance advisory Career Nomad Patrice Williams-Lindo (who recently shared IT strategies to ensure that AI supports a fair hiring process).

Apply now! The average job opening received about 242 applications in 2025, according to Greenhouse data shared with Business Insider. That’s nearly triple the number of applicants per job in 2017, which had a similar national unemployment rate.

Erin Bortz, manager of corporate recruiting at cybersecurity company Huntress, receives anywhere between 600 and over 2,000 job applications per role these days. To handle the sea of CVs, AI capabilities are essential for quick responses to top candidates.

“If it takes us two weeks to get back to them and say, ‘Hey, we are interested in talking to you,’ they may have already accepted another offer at that point, so we could be missing really good candidates because we’re not moving fast enough,” Bortz said.

Since September 2025, the company has relied on Endorsed, an AI tool with a fraud detection layer, to review applicants—a critical feature as new-hire imposters emerge. Even with that tool in place, though, a human in the loop remains critical for review.

The tool, for example, could pull suggested keyword criteria from a marketing job description. It takes a human reviewer like Bortz to determine if the AI is “reading the right signals.” From there, Bortz might have to guide the search with additional criteria—technical search engine optimization (SEO) experience, for example—and rereview what comes back.

“No matter what these tools are telling us or what we’re telling them to get the result, we are ultimately responsible for the folks that are coming through and [knowing] the outcome of the tool that we’re using,” Bortz said.

The people on the team have important roles as they use AI hiring tools, including:

• Human-in-the-loop screening (validating top and low matches)
• Looking for “soft skills” that tools can’t recognize
• Updating and maintaining job requirements to ensure correct matches are made
• Assigning the “weight” attached to each requirement within the tool, which impacts a candidate’s score

“We should be owning the ultimate decision that affects a candidate,” Bortz told us.