Leading Employer Organization Issues Report to Guide Responsible Data and Artificial Intelligence Practices in Employment
Technical Advisory Committee formed to inform responsible practices about the use of Artificial Intelligence in Employment Decision Making
The Institute for Workplace Equality (“The Institute”) announced the release of a significant report by a Technical Advisory Committee (“AI-TAC”) it sponsored to address issues about the use of Artificial Intelligence (“AI”) in employment.
The “Technical Advisory Committee Report on EEO and DEI&A Considerations in the Use of Artificial Intelligence in Employment Decision Making,” provides a series of key findings and recommendations for employers and developers to responsibly use AI-enabled tools within the current regulatory and legal framework for equal employment opportunity (EEO) and non-discrimination.
The report culminates 18 months of analysis and in-depth discussion by the members of the AI-TAC, a multi-disciplinary group of 40 experts. The group consisted of lawyers (representing both employers and workers), industrial-organizational psychologists, developers, vendors, data scientists, statisticians, and labor economists. The TAC was chaired by Victoria A. Lipnic, partner at Resolution Economics, LLC and former Commissioner and acting Chair of the U.S. Equal Employment Opportunity Commission.
HR Technology News: Key Characteristics Of An Effective Employee Management System
“This Report provides a 360-degree technical review designed to help all stakeholders – employers, applicants and employees, developers, and regulators – better understand and navigate crucial EEO and anti-discrimination issues as they apply to the use of AI-enabled employment tools within the current regulatory framework,” Ms. Lipnic said.
The Report gives examples of uses of AI in employment and takes a deep dive into issues around transparency and fairness, data collection, the Uniform Guidelines on Employee Selection Procedures and Statistics and Adverse Impact. It provides recommendations on the proper way to create data samples to develop and evaluate an AI-enabled tool so as to avoid building in bias; benchmarks for providing appropriate notice to applicants regarding the use of AI-enabled tools; and considerations to take into account when validating AI-enabled tools.
In addition to recommendations focused on the use of AI-enabled employment tools generally, the Report notes that AI tools that use machine learning (also known as “intelligent AI”) raise distinct issues. Machine learning is a form of AI that “learns” and changes on the basis of data or experience, without being explicitly programmed. The Report recommends that employers and vendors take special care in regard to such tools. This includes:
Top HR Tech Insights: HR Technology Highlights – HR Tech Daily Round-Up For 20 December 2022
Developing a policy for version control and associated guardrails around substantive changes to the functioning of the machine-learning tool.
Ensuring that the data used inmachine learning as well as the key choices made in collecting, cleaning, training, and maintaining a dataset and building a selection algorithm should be documented and a rationale provided. Documentation should be sufficient to assess data reliability and validity and allow for computational reproducibility.
“Over the past few years, we have been counseling employers who are increasingly using AI-enabled tools in their employee life cycle. They want to do so on the right side of compliance. We organized this technical advisory committee to better inform them how to do so,” said David Fortney, co-founder of The Institute for Workplace Equality and Shareholder at Fortney & Scott, LLC.
David Cohen, co-founder of The Institute and President of DCI Consulting added: “There are many important subtleties associated with standards for employee selection and those should be applied to AI-tools, just as they are to traditional tools used by employers. We wanted this TAC to raise that awareness.”
HR Tech Insights: How AI Candidate Matching Helps Recruiters Be More Productive
[To share your insights with us, please write to sghosh@martechseries.com]