The Aftermath of NYC AI Law on Employment Decision Tools

Since the arrival of AI and Chatgpt, the face of the world has been changing rapidly, every department of the corporate world is directly or indirectly related to AI and its tools. Especially in hiring and recruitment, AI tends to be most beneficial. Most companies use Automated Employment Decision Tools (AEDT) to hire and train new employees. 

NYC AI Law requires companies to use A.I. software in hiring to notify candidates that an automated system is being used. It has been making legislation since 2021 to create a better work environment in the corporate world. It has proposed many laws and regulations that AEDT can use to increase efficiency in hiring procedures. 

The directions were to enforce in April 2023, but after many changes and new rulings, it is about to be implemented on July 5th, 2023.

What does AEDT mean in NYC AI Law?

NYC AI law 144 defines AEDTs 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.”

It means if a company decides to hire a person, the tool can identify the most suitable candidate, call him for interviews and training, and substantially hire them for the position. 

Through this, Law and AI go side by side to provide better results. It cannot replace the human need as the machine cannot asses human capabilities and skills. But it gives a clear path and pattern of how a company hires a person in a specific group.

What Does the New Law Say?

According to the new law, all employers must follow four instructions before hiring a person. 

1. Suppose they use AI or AEDT for hiring purposes. They must commence an independent bias audit of the tool. 
2. After a detailed audit, an employer must publish the audit report publicly. The notice must mention the mode of AEDT, the date of AEDT, and the number of applicants in that specific process time. 
3. They must notify all the employees and applicants about using the tool. They can send an email, a message, via newspaper, or any other source. It is also mentioned that the information on AEDT must be in the applicant’s knowledge before ten business days of the interview or proceedings. 
4. They should also mention if any applicant asks them for an alternative mode of hiring or not. 

In Case of Violation

According to civil law, any violations of the AEDT law will cost the employer $500 per day. After that, the fine chargers are $500 to $1500 per day with every violation. It applies to all the business corporates and candidates in New York City. 

Pros of AI laws

Right after the notification, there is a discussion about whether the new law is suitable for the future of AI or not. It is discussed in the following points. 

• Efficient: With new laws and legislation, the hiring and recruitment process will be transparent like glass. It will give a clear pattern for hiring an employee. It may not support any specific race, color, or nationality. It opens a lot of new doors for the enterprise. 
• Cost Saving: It will save time and recruitment costs. Humans are slow in collecting data about specific skills, but a machine can do it in minutes. It will help you make decisions about hiring or promoting an employee. 
• Data-Driven Insights: With AEDT and audited bias reports, the machine has every applicant detail. So with the help of human intelligence, companies can make better decisions in less time. 
• Availability: One of the most significant benefits of this law is that it is applicable 24/7. Since the machine and recorded data are involved, anyone from the company can access it and get the required information in a few clicks. 

Cons of Law and AI

With every perk, there are consequences. If humankind is getting a lot of benefits out of it, it may also lack in some departments. Some of the cons are as follows.

• Algorithmic bias: With the new law, the probability of this flaw is lower. But it must be addressed. If data has input into the machine, the device can give biased results. If some employer hires from a specific group of people, the engine will prioritize that list. Another chance is that if you provide a list of 100 people and the priority list is the top 20, and the machine will not read the skills of the other 80. Sometimes it can make such decisions that can create more chaos and confusion. 
• Lack of human judgment: You must acknowledge the power and intelligence of human judgment. When the machine is more involved, humans can leave things on automated systems, which may affect results. In recruitment, a human can judge applicants’ skills, talents, and abilities. AEDT can affect human judgment. 
• Security and Privacy: With all the reports and data published online, it is challenging to maintain the safety of the company’s credentials. A whole new system must be developed to save the company’s data in the future for sure. 
• Over-reliance on Quantifiable Metrics: As we discussed above, this law can make machines more powerful. It will gather all kinds of information about the company and applicants. With time we can overlook essential qualities that are harder to measure, such as creativity, leadership ability, or adaptability.