As companies continue to rely on technology in their operations, more organizations are using artificial intelligence (AI) tools to screen job applicants, manage employee performance, determine promotions, and set employment terms and conditions. While these tools offer the promise of significant time and cost savings, there is a growing concern about their impact on equal employment opportunity (EEO). In response, some cities and states have enacted laws to regulate the use of AI in employment decisions. One such law is New York City’s (NYC) Artificial Intelligence and Bias Task Force (AIBTF), which requires employers using AI-enhanced hiring and employment decision tools (AEDT) to obtain an independent audit of their software to ensure that it does not undermine EEO.
The definition of AI-Enhanced Hiring and Employment Decision Tools (AEDT) refers to the use of artificial intelligence technology in the recruitment and hiring process. These tools leverage algorithms and data analysis to screen job applicants, evaluate their qualifications and skills, and predict their success in a particular role. AI-enhanced hiring and employment decision tools aim to streamline the hiring process, reduce biases, and improve the quality of hires while optimizing time and cost.
Under NYC law, AEDT refers to any tool that employs AI to assist or replace discretionary decision-making by employers. The law defines AI broadly, including machine learning, expert systems, natural language processing, and neural networks. While AEDT can identify candidates from large pools of potential job seekers quickly, it can also produce biased results if not implemented correctly.
AEDT presents some unique challenges that employers need to consider. While AI tools can save time and effort in identifying qualified candidates for a job position, these tools can also discriminate based on race, gender, age, national origin, and other EEO-protected classes. Additionally, there is a risk of requesting inappropriate information or medical record releases that could be considered discriminatory or stigmatizing.
NYC’s AEDT law requires employers to obtain an independent audit of their AI tools within one year of use and make the audit results publicly available on their website. Additionally, employers must provide notification to applicants and employees before using AEDT, informing them of the process for requesting an alternative selection process or reasonable accommodation.
To ensure that employees and job applicants are aware of AEDT use, the NYC law requires employers to provide a notice that ensures employees have ten days before AEDT use begins to request an alternative selection process or reasonable accommodation.
Employers must hire independent auditors to review AI software systems under the NYC law. These auditors are defined as individuals or groups capable of providing objective and impartial judgments regarding bias audits of AEDT.
The NYC law has far-reaching implications, not just within its jurisdiction. Recent guidance from the Equal Employment Opportunity Commission extends extraterritorial liability to employers across the country for violations of equal employment laws that arise concerning AI software. Therefore, it is increasingly essential for all companies to ensure compliance by conducting bias audits.
AI can enhance recruitment efficiency, but it must be used judiciously to avoid violating equal employment laws. Employers are responsible for ensuring that their tools are free from bias and ethical considerations. The AEDT law in NYC confirms that the law is still catching up with new approaches in recruitment technology. Nonetheless, it exemplifies the shift that is taking place in the way AI recruitment is governed. Companies should consider embracing the new era of governance for recruitment tools that champion ethical use and aim to prevent harm, both in the immediate and long term. Employers who observe it will thrive in an era of transparency, culture, and responsibility.