Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor!

Artificial intelligence in recruiting: Is it the right tool for the job?

Efficiency gains for employers can be impressive, but EPL risks remain

By Lynette Lyngaas | Associate Vice President, Public Employment Practices Liability, Nationwide

Artificial intelligence is getting a lot of attention lately, and much of that attention is negative – focusing on the downsides and unintended consequences of AI in various industries. The reality is, AI offers benefits as well as risks, and that is especially true when this emerging technology is used in recruiting employees.

According to a 2022 survey conducted by the Society for Human Resource Management (SHRM), 79% of businesses worldwide are using AI and/or automation in recruitment and hiring. Of those, 85% of employers said using AI and/or automation saves time or increases effiency.

AI can be beneficial to employers as an efficiency tool, performing a variety of functions that otherwise would take human employees a long time to complete. These functions where AI is bringing speed and efficiency include:

Writing job descriptions. AI can assist in creating more accurate job descriptions and role requirements by scanning thousands of others posted by different employers.

Candidate screening. AI can scan talent databases and social media for skills that match the requested skills for specific jobs. For this reason, efficiency is one of the major advantages that AI offers, helping employers quickly identify qualified candidates and match them with available jobs.

Answering questions. Chatbots powered by AI can assist job applicants by answering questions, providing links to information, tracking the application process and re-engaging with applicants. This can help make the employer appear more responsive and engaged in the process, which is positive for job applicants.

Augmenting staff capacity. One-on-one interviews, whether conducted in person or via videoconferencing, can be time-consuming for HR staff. By identifying and narrowing a set of applicants to the job requirements, AI can free HR teams to focus on higher-level or more complex tasks.

These are all helpful at increasing employers’ efficiency, particularly at a time when many employers are looking to fill open roles. According to the Department of Labor, the number of job openings edged upward at the end of April, to 10.1 million, while hirings remained little changed at 6.1 million. That means millions of jobs remain unfilled, month after month, so far in 2023.

Liability risks in AI

As promising as AI and machine learning (ML) tools seem for recruitment and hiring, employers should remain aware of the risks that this technology can introduce. Employment practices liability (EPL) exposures are considerable for employers of all sizes. SHRM notes there are no current federal laws regulating the use of AI in employment decisions, but several existing laws apply – including Title VII of the Civil Rights Act, the Americans with Disabilities Act and Age Discrimination in Employment Act. Responsibility for ensuring compliance with these laws rests with the employer, whether it uses AI or not.

Theoretically, AI and ML used to screen candidates should be free of biases that a human interviewer might have. But, the ability of AI to analyze not only print data but also visual images, facial expressions and even voices opens the door to potential biases. Making sure biases do not creep into and influence employment decisions comes down to careful development of the algorithms on which AI and ML rely.

AI vendors may or may not be as focused on this aspect, so employers should be prepared to monitor and revise algorithms when using AI in recruiting and hiring. Systematic errors and embedded biases, such as penalizing a candidate with gaps on his or her resume, could result in unintended consequences and trigger an EPL claim.

Similarly, hiring processes and technology tools should be reviewed to ensure they do not pose undue barriers or make unauthorized inquiries of applicants with disabilities. For example, the results of a video interview may appear different for candidates with visual or hearing impairments. AI analysis of body language and facial expression should not penalize such candidates. The ADA requires employers to make reasonable accommodations for people with disabilities so they can demonstrate their qualifications for a given job. Employers and tech vendors are in the process of understanding how to offer such accommodations when it comes to AI. An opt-out for applicants to interact with human HR staff may be important for employers to consider, and making that option clear and easy to exercise for disabled candidates.