Use of Artificial Intelligence and Face Recognition for Identification of Candidates in Interviews

A software has been developed by a US company HireVue which analyses the language and tone of the candidates while also recording the visual expressions in order to identify candidates on video. The software claims that it can increase the quality f employees a firm hire by 88 percent while also bringing diversity by 55 percent. This can lead to an increase in the filling of roles. As a result more than 700 companies have already started adapting this method all across the globe. Some of the major companies such as Urban Outfitters, Unilever, Vodafone has already conducted 100,000 interviews in the United Kingdom till date.

The software makes use of algorithms in order to evaluate an applicant in a 15 minute video against 25000 data which are collected from several of successful previously hired applicants. A professor of human-computer interaction at a university in London says that this method may favous those people who are good at giving video interviews while proving to be disadvantageous for those who would have actually been good for the job simply because they are not goof at a video interview.

Several other reviews are coming from other people. One such review coming from Griff Ferris who is the Legal and Policy Officer for Big Brother Watch. He says that this method could be an extrenly chilling experience for several job seekrs. He also mentions that the software could be biased in some way, until and unless it has been trained on highly diverse dataset. It could be discriminating to candidates coming from several certain background who are actually good for a job. On the other hand Loren Larsen who is the chief executive of this software says that he would prefer to be screened by an algorithm which treats him fairly instead of a system where a job depends on how tired and exhausted the recruiter is.

News Source: Daily Mail

Leave a Reply

Your email address will not be published. Required fields are marked *