New Evidence Rules and Artificial Intelligence

Download a printable PDF of this article.

Imagine this situation: Jane Jones applied for a senior programmer vacancy advertised by Digital Solutions Inc. Her résumé boasts undergraduate and graduate degrees in computer science, 15 years of successful programming experience in a series of increasingly demanding programming jobs, and annual job performance evaluations of “always exceeds expectations” or better.

The job went to Bill “Slacker” Bailey instead of Jane. His résumé shows that he took various computer-related courses for two years at a community college, but never obtained an associate degree, and gained five years of job experience as a tech-support specialist at a local big-box electronics store and another 10 years as the founder of “HakkersPlace,” an online blog devoted to computers and programming.

Jane filed a Title VII gender discrimination case against Digital in federal court. The essence of her claim is that she is demonstrably better qualified than Slacker and that by giving the job to him, Digital discriminated against her on the basis of her gender.

Digital vehemently denies that discrimination played a role in Slacker’s selection and asserts that despite her more impressive résumé, Jane was the lesser-qualified candidate. In support of its defense, Digital disclosed during discovery that it retains a company by the name of WorkerMatch to help it find the best candidate for each of its job openings. According to Digital, WorkerMatch has developed a computer analytics method of identifying the most qualified applicants for any particular job vacancy.

Here is how it works: WorkerMatch asks Digital to identify the top-performing employees doing the same type of work as is required for the job vacancy. Then, for a time, it digitally monitors how they perform their jobs. Next, the employees participate in a far-ranging online interview about not only how they do their jobs, but also their interests, hobbies, use of language, and mannerisms.

Then, using a computer program that incorporates an artificial intelligence (AI) algorithm, WorkerMatch analyzes the recordings of the high-performing Digital employees and develops an online questionnaire tailored to the needs of the particular Digital job vacancy. All job applicants must complete the questionnaire; those who score well then participate in an online video-recorded interview.

Again using the AI algorithm software, WorkerMatch compares the applicants’ questionnaire responses and online interview results against the results of questionnaire responses and recorded interviews of Digital’s top performers. The software then prepares a rank-ordered list of the 10 most qualified candidates.