SHRM Research: AI Use on the Rise, Ethics Questions Remain

​Most talent acquisition professionals who use artificial intelligence and automation technology say it improves time-to-fill, but they still have some uncertainty about whether the tools protect against bias.

That's according to recently released research from the Society for Human Resource Management (SHRM). The survey was conducted in February among 1,688 SHRM members representing organizations of all sizes in a wide variety of industries across the United States.

The research shows that about 25 percent of organizations report using AI—or more accurately intelligent automation—to support HR-related activities, mostly for recruitment and hiring, followed by learning and development, and performance management.

Of those who use these tools, 69 percent say the time it takes to fill open positions is somewhat or much better due to their use of the technologies. However, only 40 percent of employers say their technology vendor is very transparent about the steps taken to ensure the tools protect against bias.

"The No. 1 issue in using AI for recruitment is to ensure it is being used in a responsible manner and then being able to prove this," said Jeanne Meister, executive vice president with Executive Networks, a San Francisco-based resource group for HR leaders, and the founder of the Future Workplace Academy, which provides learning and research on the future of work. "AI must be transparent, explainable and fair."

Over half (54 percent) of respondents who use automation or AI to support HR activities said they've faced challenges when using the tools, including not having the resources to properly audit or correct AI algorithms and the algorithms inadvertently overlooking or excluding qualified applicants or employees.

"Clearly, automation helps HR deliver value to organizations, especially when it comes to acquiring top talent," said SHRM Chief of Staff and Head of Government Affairs Emily M. Dickens. "But we need to be assured the tools we use do not lead to bias in the hiring process, performance management or other areas of HR."

While 30 percent of respondents said the use of automation or AI improves their ability to reduce potential bias in hiring decisions, 46 percent would like to see more information or resources on how to identify any potential bias when using these tools.

Dickens said SHRM provided the research to the Equal Employment Opportunity Commission (EEOC), the Department of Labor and members of Congress to help inform future policymaking.

"I am pleased to note the EEOC launched an agencywide initiative to ensure AI systems comply with civil rights law," she said. "We look forward to being a partner in these important efforts."

'Safety Checks'

Meister said companies that work with AI technology vendors should require these vendors to audit the data on a regular basis, "like safety checks on automobiles," and to be transparent about the data behind the algorithms. "This level of transparency is already being required in the consumer area," she said. "Companies should require technology vendors to share how their AI models were built, how they are trained, the demographics of those training the algorithms and how often they will provide this information to their clients."

She added that using AI in HR is a change management process and companies need to assemble a cross-functional team of legal, finance, business unit and HR leaders to continually review the ethical usage of AI—not just the metrics of increasing the speed to hiring.

"Companies can do this by creating governance around the use cases of AI in the organization and giving this governing body the responsibility of overseeing ethical AI."

Automation Boosts Efficiency

AI and automation help companies save time and/or increase efficiency (cited by 85 percent of respondents), automatically filter out unqualified applicants (64 percent), and improve their ability to identify top candidates (44 percent).

About 68 percent of HR professionals said the quantity of applications they must manually review is better due to their use of AI or automation, while 59 percent said the quality of the applications is better.

"One of the biggest benefits of using AI for hiring is that AI is able to identify stronger candidates for a job faster than humans can," Meister said. "While speed of hire has dramatically increased using the technology, heads of talent acquisition must also focus and track the ability of AI to source higher-quality candidates."

SHRM found that small and midsize organizations are generally not using automation or AI for HR. Just 16 percent of employers with fewer than 100 workers use automation or AI compared with 42 percent of employers with 5,000 or more workers.

The research also revealed that 25 percent of organizations plan to start using or increase their use of automation or AI in recruitment and hiring in the next five years, and 20 percent of organizations plan to start using or increase their use of those tools for performance management over the next five years.