According to IBM CEO Ginni Rometty, IBM AI tech can predict which employees in the U.S. workforce are looking for new jobs with 95% accuracy.

The recently patented “predictive attrition program” was developed alongside IBM’s Watson to make predictions about employee flight risk as well as recommend solutions for employers to better engage those employees. The success of the program, according to Rometty, was achieved through an analysis of several different undisclosed data points, reportedly resulting in the 95% rate of accuracy.

"It took time to convince company management it was accurate," Rometty said, yet the AI has so far saved IBM roughly $300 million in retention costs, she claimed.

The predictive attrition program is just one component of an entire suite of AI-based tools IBM has developed to disrupt the traditional human resources (HR) industry, according to Rometty. Using AI tools that look for data patterns to assess an individual’s skills and identify future work opportunities, as well as AI tools that replace human performance evaluations, have all led to a 30% reduction in IBM’s global HR department, according to Rometty.

Calling the current iteration of HR "broken," Rometty believes that AI is better than humans at HR and, consequently, it is on the path to replacing humans in the HR industry.