These FAQs provide some background on HackerRank’s approach to the New York City Local Law related to employer usage of Automated Employment Decision Tool (“AEDT”) as they relate to HackerRank customers using HackerRank services that make use of artificial intelligence or machine learning.
What is an AEDT?
The NYC Law defines an AEDT as “any computational process, derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issues simplified output, including a score, classification, or recommendation, that is used to substantially assist or replace discretionary decision making for making employment decisions that impact natural persons.”
Are HackerRank services AEDTs covered by the NYC Law?
HackerRank does not view its Screen and Interview services as AEDTs because these services are not derived using artificial intelligence, machine learning, or similar technologies (“AI”). Nor is the output generated by these services the result of AI processing.
Plagiarism Detection/Image Analysis
HackerRank’s Plagiarism Detection and Image Analysis (our “AI Tools”) are add-on functionality that HackerRank customers can choose to enable within their use of the overall HackerRank services. These AI Tools are derived from AI and issue a simplified scoring output.
Whether these AI Tools are deemed AEDTs under the NYC Law likely depends on how the customer uses them, i.e., whether the customer uses HackerRank’s AI Tools to substantially assist or replace discretionary decision-making.
Access the Test Integrity settings of your Test to make appropriate changes based on your requirements.
The option is available under the Proctoring settings in the Test Integrity tab of a Test.
If a HackerRank customer uses the results of our Automated Tools to “substantially assist or replace” human decision-making by, for example, automatically rejecting candidates for whom plagiarism or image anomalies are flagged, then our Automated Tools will likely be considered AEDTs. But, if the customer uses the results as one of many factors in determining a candidate's fitness, they likely won’t be considered AEDTs.
What does the NYC Law require when using AEDTs?
If a HackerRank customer uses our AI Tools as AEDTs, the NYC Law requires the customer to provide New York City candidates subject to the AEDT with certain notices and the option to request an alternative selection process or accommodation.
The NYC Law also requires that our AI Tools undergo a bias audit and that the customer make that bias audit available to its candidates.
Such a tool has been the subject of a bias audit conducted no more than one year prior to the
use of such tool; and a summary of the results of the most recent bias audit of such tool as well as the distribution date of the tool to which such audit applies has been made publicly available on the website of the employer or employment agency prior to the use of such tool.
How is HackerRank assisting customers with their NYC Law compliance obligations?
If a customer uses our Automated Tools as AEDTs, HackerRank is taking multiple steps to make it easier for our customers to comply with their obligations under the NYC Law.
- Defaulted “Off.” Our Automated Tools are defaulted to “off” within the HackerRank product. Automated Tools will not be enabled unless the customer turns them on. This gives the customer a choice as to whether they want to use our tools that could qualify AEDTs.
- Customer Reminder. When a customer user turns on our Automated Tools, the user will receive a pop-up reminding the user that these automated tools use automated processing based on AI and that the customer may need to comply with applicable laws related to automated processing.
- Consent. If the customer enables an Automated Tool, HackerRank will secure the candidate's consent prior to the use of an Automated Tool. If the candidate does not give consent, the candidate will not be able to participate in the feature using the Automated Tool. This consent is not intended to replace the 10-day advanced notice and alternative method request that the customer must offer the candidate but is instead an added precaution (based on GDPR, not necessarily the NYC Law).
- Bias Audit. Our Automated Tools have gone through a bias audit. For this first year, we are using test data for our bias audit. In future years, customers will likely need to provide us with their historical data to conduct the bias audit.
- HackerRank's AI-powered plagiarism solutions have gone through an independent bias audit consistent with the NYC AI Law.