Starting April 15th 2023, New York city is expected to regulate the use of automated employment decision tools (AEDT). Specifically, employers are expected to conduct bias audits on AEDT, (both home grown and vendor sourced) through an independent algorithmic auditor. In addition, employers need to provide notices about usage of AEDT to employees or job candidates. [1] provides more details about the upcoming law. While regulatory efforts on use of AI and automation is slowing picking up traction in the US, this law is one of the first that mandates employers to carry annual audits leading to a lot of discussion on what the expectations are. In this master class, we seek to discuss the published law and bring about clarity on expectations both from a technical and legal perspective.
In Part 1 of this master class, Sri Krishnamurthy will provide an overview of the NYC bias audit law and discuss the technical specifics of what is expected in an audit.
Through examples, we will discuss:
What a bias audit entails
Terms such as Scoring rate, Selection rate, Impact ratio, Simplified output
Frequently asked questions
In Part 2 of this master class, Susie Cirilli will provide a legal perspective on the NYC audit and discuss:
The proposed regulations that have been issued by the Department. These draft regulations provide valuable information for what employers should be prepared for come April 15, 2023.
Finally, we will share details of the upcoming QuantUniversity course:
Algorithmic Auditing : Preparing for the NYC Bias Audit course
Reference
REGISTER HERE
