AI is everywhere and making an immense impact in industries like Energy, Healthcare, Manufacturing and so many more. But have you ever associated AI with Talent Acquisition? Find out what our CEO Chris Carosella has to say about how AI is improving the search for highly qualified candidates and making the hiring process a win-win for both job seekers and companies.
3 Ways to Improve Efficiency and Curb Bias With AI
By Chris Carosella
Originally published on hr.com
Screening applicants remains the biggest challenge in talent acquisition. At the same time that corporate recruiters are diversifying hiring pools, they are also being asked to do more with less. More than half of talent acquisition leaders expect to hire more employees this year than last year, but 66% say their recruiting teams won’t grow accordingly.
To solve this, corporate recruiters are turning to artificial intelligence. Because AI enables target searches by job title, industry, location, salary, education, and more, it can work much more efficiently to sift candidates, assess résumés, make contact, and even conduct initial interviews.
Dropping Costs, Gaining Time, Losing Bias Using AI in hiring can dramatically increase diversity and cost efficiency. In the three years since Unilever began using AI in recruiting, for example, the average hiring period decreased
from four months to just one. Not only that, but the time recruiters spent reviewing applications dropped by 75%, saving Unilever more than 70,000 human-hours.
The ways companies use AI vary from organization to organization. International accounting firm EY hires about 65,000 people each year and utilizes its popular AI chatbot, “Goldie,” to
implement AI programs that assist HR in screening candidates. But for companies that don’t have their own in-house AI programs, tools like
AllyO,
TextRecruit, and
Montage allow recruiters to assign AI the preliminary work and present the best candidates to HR.
Gamifying the hiring process is an ongoing trend among recruiters, too — one that
AI can facilitate. One reason Unilever has seen such impressive results is that it asks candidates to play a selection of games to test their aptitude, logic, reasoning, and willingness for risk. AI then assesses the candidates’ suitability for the positions they have applied for by comparing their results against those of successful predecessors.
Some companies even use AI to assess candidates’ word choices, facial expressions, and speech patterns to determine how they will fit with the culture. But there’s a caveat here: Using AI to determine cultural fit could result in a hiring process that’s just as biased as humans. For example, Amazon’s AI creation turned out to be
biased in favor of men when it came to hiring tech talent. It turns out that their algorithms had been taught to look for résumé terms that mirrored those of past job applicants who, unsurprisingly, had been primarily male.
Despite the missteps we’ve seen, though, AI has the potential to reduce or even eliminate human bias in corporate recruitment. When AI software uses predictive analytics to calculate a candidate’s likely success in a role, it can give recruiters and managers the power to make data-driven hiring decisions. The trick is to start with the right parameters. Begin by using AI to find the pattern of words that works to get engagement from different people, and then use a tool like
Textio to identify words or phrases that could alienate certain groups.
Strategies for AI Integration As AI continues to improve, start embracing it by taking these steps to build a solid foundation:
1. Clarify your needs. Before inventing elaborate applications and investing time and money in developing custom software programs, think through exactly what the company wants and needs from AI. What are the specific cases in which AI could solve problems or provide value? Clear goals and use cases will get crucial buy-in from everyone involved, from the C-suite on down.
2. Diversify the team. AI is built by humans, and the humans building AI are part of the tech industry, a sector
notably lacking in diversity. Of the 641 people working on Google’s “machine intelligence” efforts, for example, only 10% are women, while only 22% of Facebook’s technical workers are women. To avoid building bias into your algorithms, tap a team of men and women with diverse tenure and types of experience to lead the project. Just because AI is technical doesn’t mean the tech people should run it. Instead, AI needs vast and varied input to be successful.
3. Look to the future. Any recruitment initiative needs to consider the company’s needs not only right now, but also in the long term. As you work to fill your immediate openings, think strategically about what skills and experience will be needed in future positions. You can then be intentional about determining the search terms and parameters that find candidates who will grow with the business.
Whether simple or complex, AI takes its orders from humans. If we take steps to feed it good information, AI can improve outcomes for job seekers and companies alike.