Max Wesman, chief product and strategy officer at GoodHire, an employment screening company, says machine learning and AI are key in helping HR professionals and recruiters more efficiently process background checks for large volumes of candidates.

Wesman says GoodHire has used AI to help HR professionals determine the type of information they want to exclude from their decision-making process — such as minor criminal offenses — and automate the process for reviewing candidates by marking those who meet requirements or need additional review.

“We’ve seen this use of AI particularly benefit employers who hire in large batches, including at hiring fairs, where they need to review and onboard new employees quickly,” Wesman says.

Dave Marko, managing director at Acumen Solutions, says tools powered by AI and predictive analytics are helping companies shift from trailing indicators for candidates to leading indicators. Trailing indicators, he says, give information to recruiters when it's too late to make an impact on the tenure and performance of candidates. Leading indicators, on the other hand, help recruiters get ahead of the curve and understand key workforce metrics, even before a candidate is hired.

“Employers know that bad hires can be very expensive and are now taking a more measured approach to vetting candidates,” Marko says. “These data points essentially allow empower recruiters to target specific skills and attributes that lead to higher-performing, longer-tenured employees.”

These tools are not without challenges, however. Robin Schwartz, HR director at Career Igniter, says that while AI is improving recruiting processes and potentially removing unconscious bias, it may also remove an important element of the hiring process: emotional intelligence.

For example, Schwartz says AI platforms may be able to cross-reference job posts with all incoming resumes and find the top two applicants, but often it can’t effectively identity red flags like someone who has changed jobs four times in two years. “While AI is making positive impacts on recruitment, it’s also potentially removing some of the best applicants out there,” she says.

Aman Brar, CEO of Canvas, a text-based interviewing platform, says more companies are looking for recruiting methods that align with the communication preferences of today’s talent. He says organizations are also increasingly worried that traditional recruiting techniques could cause them to lose out on worthy mobile-first applicants.

As a result, more companies are switching to text-based recruiting and AI-powered chatbots, which leverage algorithms that can generate pre-built interview questions to increase applicant quality and reduce time to fill, Brar says.

Matos, of Paradox, which produces an AI recruiting assistant called Olivia, says it comes down to improving the overall candidate experience with better interactions. “Candidates want and expect information quickly, easily and specific to them,” he says. “They don’t want to read an FAQ page on a website. They want to ask someone their specific question and receive an answer, at that moment. Tasks like these are a great fit for AI, who can perform these high-volume, repetitive tasks that are time consuming for recruiters, but valued by candidates.”

Early AI models aimed to accurately reproduce what a recruiting professional would do, such as parsing through resumes to decide which candidates should be interviewed, says Gabriel Bianconi, founder of Scalar Research, a full-service artificial intelligence and advanced analytics consulting firm. “The problem is that HR professionals have biases in this selection process — albeit usually unconsciously — so the models ended up being biased as well,” he says. “In fact, the models identified those implicit biases and reinforced them, making them even more harmful.”

Bianconi says newer AI models seek to solve for this problem, although this can be hard to achieve because there are subtle correlations between different groups and the data associated with them, such as the fact that men and women tend to use slightly different vocabulary.

Some of these new approaches may already be working when it comes to diversity. Matt Dodgson, one of the founders of Market Recruitment, a U.K.-based recruiting company focused on helping B2B and tech companies find marketers, says cites success with the AI offering Textio to generate a more diverse candidate pool.

“This solution has made our job postings so much more successful,” Dodgson says. “With its incredible database, practically all gender bias has been removed from our job postings and we can see what language might work better on certain audiences. We now receive almost twice the amount of female candidates compared to when we crafted the job postings ourselves.”