Final NYC Rules on the Use of Automated Employment Decision Tools Published – Enforcement Delayed until July 5, 2023
April 25, 2023
In a Nutshell
|
What Happened |
The rules to accompany New York City Local Law 2021/144 have been finalized. The rules will take effect on May 6, 2023 and be enforceable on July 5, 2023. A summary of the requirements under the finalized rules are outlined below. |
|
Who Does this Affect |
Employers or employment agencies who use automated employment decision tools for employment decision-making purposes in New York City. |
|
How this Affects You |
Violators of the law can face penalties ranging from $500.00 to $1,500.00 per violation. A company will be subject to a separate violation for each day that it is out of compliance with Law 2021/144. |
|
What You Can Do Now |
See guidance below entitled “What Actions Must Employers and Employment Agencies Take When Using AEDTs?” |
Summary of Local Law 2021/144 and Accompanying Final Rules
What Happened?
On April 6, 2023, after several rounds of public comment and revisions, New York City finalized the accompanying rules (“Rules”) to Local Law 2021/144 on Automated Employment Decision Tools (“Law”). These Rules will come into effect on May 6, 2023, and the rules will be enforceable as of July 5, 2023. For a general overview of the risks and benefits to using AI, please see our client alert: A Double-Edged Sword: The Benefits and Risks of AI in Business.
What is Local Law 2021/144?
The Law imposes restrictions on an employer’s or employment agency’s use of automated employment decision tools (“AEDTs”) to make employment-based decisions[1] pertaining to candidates and/or employees residing in New York City. The Rules provide clarity to businesses on what is an AEDT and how to comply with the Law.
What is an AEDT?
AEDTs are broadly defined as “any computational process, derived from machine learning, statistical modeling, data analytics, or artificial intelligence[2], that issues [a] simplified output[3], including a score, classification, or recommendation, that is used to substantially assist or replace discretionary decision making[4] for making employment decisions that impact natural persons.” Tools that do not materially impact natural persons (e.g., a junk email filter, firewall, antivirus software, calculator, spreadsheet, database, data set, or other compilation of data) are not considered AEDTs.
What Actions Must Employers and Employment Agencies Take When Using AEDTs?
Employers and employment agencies may only use AEDTs to make employment-based decisions if it complies with the three requirements described below.
- Complete a Bias Audit and Post the Results on its Website.
The AEDT must be subject to a bias audit by an independent auditor within one year of first using the tool and on an annual basis thereafter. A summary of the audit results must be posted on the company’s public facing website for at six months after last using an AEDT. Further details on what the bias audit must cover (including the data types that should be used to conduct the audit) as well as what information must be posted on the company’s website is covered in the Rules.
- Provide Notice / Information to Candidates / Employees.
Employers and employment agencies who use AEDTs to make employment-based decisions must notify affected candidates or employees ten days prior to using the AEDT of the following: (i) that an AEDT will be used in connection with assessing the individual for a prospective position or promotion; (ii) the job qualifications and characteristics the AEDT will use to assess the individual; and (iii) instructions on how the individual can request alternative assessment mechanisms or accommodations. Within thirty days of a written request from an affected individual, the employer or employment agency must also provide to that individual: (i) the type of data collected for the AEDT, (ii) the source of such data, and (iii) the company’s data retention policy. The Rules provide further guidance on how to properly provide notice to affected individuals.
- Provide Affected Individuals an Alternative Assessment Mechanism or Accommodation.
Lastly, employers and employment agencies must provide an alternative selection process or accommodation if requested by the affected individual.
How We Can Help
[1] This term is defined under the Law as screening candidates for employment or employees for a promotion.
[2] This term is defined under the Rules as “a group of mathematical, computer-based techniques: (i) that generate a prediction, meaning an expected outcome for an observation, such as an assessment of a candidate’s fit or likelihood of success, or that generate a classification, meaning an assignment of an observation to a group, such as categorizations based on skill sets or aptitude; and (ii) for which a computer at least in part identifies the inputs, the relative importance placed on those inputs, and, if applicable, other parameters for the models in order to improve the accuracy of the prediction or classification.
[3] This term is defined under the Rules as “a prediction or classification as specified in the definition for ‘machine learning, statistical modelling, data analytics, or artificial intelligence’. A simplified output may take the form of a score (e.g., rating a candidate’s estimated technical skills), tag or categorization (e.g., categorizing a candidate’s resume based on key words, assigning a skill or trait to a candidate), recommendation (e.g., whether a candidate should be given an interview), or ranking (e.g., arranging a list of candidates based on how well their cover letters match the job description). It does not refer to the output from analytical tools that translate or transcribe existing text, e.g., convert a resume from a PDF or transcribe a video or audio interview.”
[4] Under the Rules, this term means: “(i) to rely solely on a simplified output (score, tag, classification, ranking, etc.), with no other factors considered; or (ii) to use a simplified output as one of a set of criteria where the simplified output is weighted more than any other criterion in the set; or (iii) to use a simplified output to overrule conclusions derived from other factors including human decision-making.”
Click here to view a downloadable PDF of the legal update.
This Cybersecurity, Data Protection, & Privacy Alert is intended to keep readers current on developments in the law and is not intended to be legal advice. If you have any questions, please contact Matthew H. Meade at 412.566.6983 or mmeade@eckertseamans.com, Elizabeth Wilson at 215.851.8497 or ewilson@eckertseamans.com, a member of our Cybersecurity, Data Protection, & Privacy Practice Group, or any other attorney at Eckert Seamans with whom you have been working for further information and assistance.
Share This Post
Authors
