Artificial Intelligence and the next wave of employment litigation in New York

Artificial Intelligence (“AI”) can be a valuable tool for employers searching for the next “perfect” candidate or assessing employee performance. However, as employers rely more heavily on AI-driven technology in hiring, promotion, and performance evaluation, courts and lawmakers are scrutinizing whether these technologies may give rise to new avenues of workplace discrimination. In New York, legal risks associated with AI are rapidly emerging as a major source of employment litigation.

Employers routinely utilize AI to review resumes, rank applicants, monitor employee productivity, and assist in making disciplinary and performance-related decisions. While this may improve efficiency and streamline decision-making processes, using AI in these ways can also raise concerns that automated decision-making may unintentionally discriminate against protected groups or individuals.

New York has been at the forefront of regulating artificial intelligence in employment. New York City, often a model for other municipalities, was among the first cities in the country to require oversight of employers’ use of Automated Employment Decision Tools (AEDTs). AEDTs are AI-powered technologies that are used in hiring or promotions to assist in making employment decisions by screening resumes, rating candidates, assessing interviews (including video interview analysis), performance evaluations, and more. Some examples include screening software that filters applications based on keywords and interview platforms that analyze candidate responses and behaviors.

Under New York City’s Local Law 144, employers that use AEDTs must conduct annual bias audits, provide notice to applicants and employees before using these tools, and make certain audit results publicly available. The audits must be conducted by an independent auditor and assess the tools’ impact on protected categories, including sex, race, ethnicity, and age. Noncompliance can result in significant daily civil penalties.

While New York City was the first to pioneer the regulation of AEDTs, other jurisdictions have since enacted or introduced similar legislation, reflecting a broader trend toward increased regulation of AI in the workplace. Employers throughout New York State should expect similar scrutiny from legislators, courts, and litigants as the use of AI continues to expand.

The primary concern driving AI-related regulation and litigation is what is known as “algorithmic bias.” AI systems learn from historical data, and if the data relied upon reflects existing or past discriminatory practices, the resulting recommendations may replicate or amplify those inequities. As a result, employers that rely on AI-driven decision-making tools may face liability if those tools disproportionately impact members of protected groups, even where discrimination is not intended.

For example, a hiring algorithm utilizing an employer’s historical hiring data may be more likely to favor applicants who share characteristics with individuals previously selected for employment. If the employer’s workforce lacked diversity, the algorithm may unintentionally disadvantage certain protected groups, including women, older workers, racial minorities, or individuals with disabilities.

As employers are increasingly incorporating AI into employment decisions, plaintiff’s attorneys are paying closer attention to whether these technologies result in disparate impact on protected classes. Discrimination claims arising from automated decision-making remain a significant area of litigation risk.

In addition, AI presents unique challenges under the Americans with Disabilities Act and the New York State Human Rights Law. Automated screening tools may inadvertently exclude qualified candidates with disabilities. For example, resume screening software may reject applicants with nontraditional work histories, or productivity monitoring may fail to account for reasonable accommodations. Because employers remain responsible for employment decisions made with the assistance of technology, reliance on a third-party AI vendor does not shield an employer from liability. Courts will focus on whether employers exercised reasonable oversight of these systems to ensure compliance with anti-discrimination laws.

Transparency has become one of the most controversial issues surrounding the use of AI in employment decisions. Plaintiff’s attorneys and employee advocacy groups are increasingly arguing that applicants and employees should be notified when AI is being used and provided with meaningful information regarding the role AEDTs play in hiring, promotion, discipline, and termination decisions. Recent lawsuits have challenged AI hiring tools on the grounds that candidates were evaluated through automated processes without their knowledge or consent. Critics contend that when employers rely on algorithms, applicants may have no meaningful way to understand why they were rejected, how to identify potential bias, or challenge inaccurate results. This can make it difficult to determine whether an employment decision was based on business factors or unintended bias.

In response to these transparency concerns, regulators have increasingly focused on disclosure and accountability. Employers should anticipate increased obligations relating to notice, disclosure, record retention, and documentation of AI-assisted decisions. Employers that are unable to explain how an algorithm operates may face significant challenges defending employment claims.

To minimize litigation risk, employers should do the following:

• Conduct regular internal audits of AI-assisted employment systems, in addition to any required external audits.
• Review vendor agreements to ensure access to information regarding how automated tools operate.
• Always maintain human supervision over employment decisions and review all hiring and evaluation tools.
• Train all human resource personnel regarding AI-related risks.
• Document the legitimate business reasons supporting employment decisions.
• Provide reasonable accommodations for individuals with disabilities, especially if automated systems may disadvantage them.
• Stay updated on evolving regulations governing AI in the workplace.

Artificial intelligence can be valuable in the workplace, but it is also creating a new wave of employment litigation. Employers must remain vigilant and understand that AI will be subjected to the same anti-discrimination laws that govern traditional hiring, promotion, discipline, and termination decisions. Being proactive today may significantly reduce the cost of future litigation. This is likely to remain one of the most closely watched areas of workplace litigation for the foreseeable future.

Jennifer A. Shoemaker is a Partner in Underberg & Kessler LLP’s Litigation, Labor & Employment, and Family Law practice groups. She represents clients on a wide variety of litigation, family law, and labor & employment issues, including divorce, custody, and support cases, day-to-day labor and employment matters, and labor and employment disputes. She can be reached at [email protected].