Employers are relying heavily on the capabilities of Artificial Intelligence to conduct multiple interview levels, which may create a bias risk
Artificial Intelligence is a powerful technology that can be applied to almost any responsibility. In a growing trend, AI is heavily being deployed to conduct the initial interview rounds. Interviewers say that this solution has helped them filter out thousands of candidates and also helps them create a character summary that matches their job description.
Artificial Intelligence is growing rapidly as the preferred interview tool for brands ranging from Quick Service Restaurants to software startups. Barb Hyman, the CEO of an Australian AI startup, Sapia.ai, said that the application process for these companies invites thousands of candidates who may never land the job. However, providing a good post-application impression is critical for these brands as most potential candidates are also future customers.
“They’re going to reject hundreds of thousands, maybe millions, but they don’t want to lose them,” she says, “and they want them to feel that they got a fair go.”
Sapia’s core offering includes conducting first interview rounds through textual Q&A sessions. The interview candidate provides short answers of 50 to 150 words for questions regarding skills and experience. The AI tool maps out the answers to form characteristics based on personality traits, communication skills, critical thinking abilities, and other aspects to form the candidates' personality profiles. Employers get a shortlist of the matching candidate profiles that filter down over thousands of resumes and candidates.
“I didn’t want to have a huge team of people just doing a really mundane screening of entry-level roles,” said Rose Phillips, head of partner resources at Starbucks Australia. “The other reason was to try and improve the quality of applicants that we were referring to our store managers, [since] they don’t have a huge amount of time to spend on recruitment.”
After looking at the capabilities, the question arises about the AI tool's efficiency in preventing biases. Top AI tools like Gemini and OpenAI’s ChatGPT have been prone to bias risk as their learning mostly happens from online scrapes of information. Even if the niche tools are trained on custom LLMs, the tools remain vulnerable to form bias based on the historical data of the candidates that they train on. AI hiring companies claim that their tools prevent bias based on gender, race, culture, and other protected classes.
However, the bots may not perform accurate judgment roles that humans can capture, as most interviews are text-based. Federal law bodies do not require hiring companies to disclose their AI usage, making visibility on the bias risk difficult. Many companies do not reveal the AI interview bot process to the candidates either, creating a gap between candidate expectations and employer hiring practices.
