An artificial intelligence simulation uses algorithms to score how fit the candidate is for their next job.
Bhavna Dabysingh and her husband are newcomers to Canada. With a foreign name on her resume, her job applications in Ottawa have been dismissed with lack of Canadian experience. People have asked her to apply for manual labour while others cannot believe why her husband is jobless because he is so overqualified. Despite over qualifications, fluency in both English and French, “we’ve seen that the bias is real….” she says.
Artificial intelligence (AI) is revolutionizing employment for people like Dabysingh who constantly struggle with unconscious bias as employers reject candidates based on their names, ethnicity, accents. Challenging unconscious bias stems from AI companies like Knockri in Toronto whose co-founder and CEO Jahanzaib Ansari struggled with biased rejection himself. AI tools are now shortlisting candidates with video interviews analyzing candidate’ responses, tonality, facial contractions or screening online resumes. AI in the workplace is an ongoing phenomenon with supporters and critics who are tentative about its capabilities. Nevertheless, 14 per cent of employers in Ontario are using AI in human resources (HR) and 16 per cent of HR professionals believe reducing bias is one of the top challenges AI can help overcome, according to a recent report by the Human Resources Professional Association. Although AI is in its early stages for employment strategies, it is making bold changes.
“We have seen a 23 per cent increase in the diversity and inclusion funnel of the shortlists,” says Knockri co-founder Faisal Ahmed.
As you look into the webcam everything from the content you speak, to where your eyes wander, to how long it takes to formulate a response, programmed algorithms-not humans are scoring if you are fit for the job.
“Humans have a lot of unconscious biases which subconsciously play a role in our decision making; especially in recruiting and it adversely affects the person that is trying to get a job,” says Ahmed, “and these biases are making a lot of good candidates fall through the cracks.”
Knockri currently shortlists candidates for companies like IBM. It uses video to assess good correlation and qualifications. The technology first scores the content of answers, then the delivery of the content (tonality), and lastly facial contractions on how candidates express their answers. They indicate attributes like confidence and communication skills. The short list is given to employers with only scores and no names or faces attached.“This technology looks into soft skills which traditional methods of resumes can’t address,” says Ahmed.
Broke and desperate for a job at the time, Ansari expressed how his struggles with unconscious bias spurred his frustrations into co-founding the company.
“As a candidate, I used to hate the fact that no one reached out to me. It was like I was stuck in the dark,” says Ansari who after some advice found a solution. “I started to anglicize my name on my resume like Jason, Jay, Jordan and honestly the same jobs that I would first try to apply to I would hear back and got a job within four to six weeks.”
A recent study suggests how specifically, applicants with Asian names have a 28 per cent reduced likelihood of getting called for an interview compared with applicants with an Anglo name even when all qualifications are equivalent and Canadian in origin. One problem is that employers are themselves disadvantaged by lacking resources to acknowledge the qualifications of applications and their different names.
And while a company could use an AI to fight bias in hiring from blind resumes, and video assessments, some critics say it’s too new of a technology to foster people’s qualifications
“There is this hocus, pocus mysterious aspect where people feel technology can do anything,” says Jeffrey Reitz, a sociology professor at the University of Toronto and co-author of the study, “Do Large Employers Treat Racial Minorities More Fairly? He suggests that AI seems hopelessly inadequate to be a summary of qualification. Reitz emphasized that most of the capabilities of AI are still unknown. Depending on the type of AI used, “when it eliminates information it also may eliminate information about qualifications,” he says.
For Reitz, it all depends on the developments of our time. For example, blind recruitment methods with orchestras assess the musical performance of performers behind a screen to reduce gender bias. Reitz feels blind recruitment is one of many ways to reduce bias and better assess qualification.
“For me, the use of AI will never replace human selection in the search for talent,” says Dabysingh. She feels a fair selection process would be asking for anonymous resumes which are similar to the blind recruitment method that Reitz discussed.
Fawad Khan develops AI to shortlist candidates based on their resume submissions and the employer’s expectations in Ottawa. “Oftentimes decisions are based on the name and the origin of the name,” he says. “And by taking the name out of the resume the focus shifts to experience and skill.”
After one submits their resume online, AI searches for basic keywords from the resume in a quarter of a second including geography, years of experience per title and position, and skills, according to Khan.
Although AI selects the top 10 candidates based on experience and skills, there is no guarantee they will get hired. AI professionals recognize that the human element is still needed even though measures are being taken to reduce unconscious bias and increase racial diversity for people like Dabysingh and her husband.
“I don’t think it is a good idea to completely eliminate the human aspect.” says Khan.”Although AI can recommend somebody, there still needs to be a human at the end to give the final call.”