NJBIA is seeking amendments on a new bill that regulates the use of automated tools in hiring decisions to minimize discrimination in employment. 

NJBIA Vice President of Government Affairs Alexis Bailey concerns are based on the fact there is “not a lot of concrete data on how, when and to what extent these different (hiring) tools are used by employers.” 

She informed the Assembly Labor Committee that NJBIA is currently partnering with a higher education institution to conduct a survey to gather more information from human resources professionals on how they are using these tools.  

“This data can then help us better understand how these tools are used before we begin regulating them,” Bailey said. 

Bailey recommended several amendments if the Legislature proceeds with the bill prior to analyzing the survey data. The first of which is to require automated decision tool companies to provide summaries of bias audits, instead of full audit results. 

“(The bill) requires companies that sell AI hiring tools to ‘provide the results of the audit to the purchaser’ once per year,” Bailey said. “We request that this section be amended to indicate that purchasers be provided with a summary of the audit results instead of the full report.  

“Requiring AI tool companies to give the entire audit report to a consumer could result in valuable intellectual property being divulged.” 

NJBIA is also seeking an amendment that allows employers to notify job candidates that they are using AI tools on initial job postings to satisfy the notice requirement. 

“This would ensure that job candidates are still informed of AI’s role in the hiring process but can provide a streamlined option for employers to provide notification,” Bailey said. 

Additionally, NJBIA is seeking clarification on whether recruitment and hiring-based websites are included in the definition of automated employment decision tool.  

“Many small employers utilize recruitment and job-posting websites to find new employees,” Bailey said. “These websites can filter candidates for employers as well as share their job listing with targeted job seekers using algorithmic technology. We suggest clarifying the definition of automated employment decision tool to exclude these types of platforms.” 

“We are thankful for the committee for hearing and considering our recommendations,” Bailey said. “We look forward to working with the sponsors on the bill.”