NJ and NY Rewire AI Hiring Practices to Sidestep Bias Pitfalls

Artificial intelligence and other similar advanced technologies are streamlining hiring, and companies are balancing between efficiency and equity.

Algorithms often carry the weight of old biases, and these systems can increase discriminatory patterns, such as in resume screening or interview evaluations, by relying on data that reflects past unfairness.

These matters illustrate the need for state-level initiatives to ensure fair practices, and employers must prioritize compliance to mitigate litigation and ensure an inclusive workplace.

New Jersey’s January guidance and New York’s pending legislation, which showcase the imperative for impact assessments and transparency notices, are not barriers to technology. They are frameworks that can support the creation of sustainable organizations.

Guidance

NJ issues AI employment guidance. New Jersey’s Division on Civil Rights issued targeted guidance that interpreted the state’s Law Against Discrimination to address algorithmic tools in employment decisions. The New Jersey Attorney General and the DCR then issued a 13-page document, applying to any automated system influencing hiring, promotions, or performance reviews that disproportionately affects protected characteristics, such as race, gender, age, or disability.

It covers intentional disparate treatment and unintentional disparate impact, with no safe harbor for “neutral” algorithms if outcomes reveal inequities. The guide provides clarity to employers on auditing and refining how they use AI, which would place companies in a leading role regarding hiring.

For practical compliance, employers operating in New Jersey and New York can perform pre-implementation bias impact assessments, evaluating training models for imbalances and effects on diverse applicant pools.

Human oversight is non-negotiable, and trained reviewers may need to intervene in AI-generated recommendations. Candidates must receive clear notices about AI’s role, including how data is used, and opportunities to request explanations or accommodations.

If companies don’t take these precautions, the consequences can be costly. Violations of the underlying LAD, now extended to AI tools through the guide, can trigger DCR investigations, with penalties under the statute ranging from $10,000 for initial violations to more than $50,000 for patterns of disregard, plus compensatory damages and injunctive relief. The financial implications of implementing the recommendations are impactful.

Being ahead pays off, and there are growing positive results from applying the steps, such as smoother applicant experiences, significant savings, and fewer red flags. For example, the $365,000 settlement in a 2023 EEOC action involving a flawed AI screening tool underscores the financial risks and led to revisions to hiring protocols. Customized certifications are being developed to assist human resource departments, but employers remain accountable.

Regulation

NY advances AI regulation. Pivoting to New York, legislative momentum is building with Senate Bill S1169A, the New York Artificial Intelligence Act, introduced in January and advancing through committees. If enacted, the bill would regulate “high-risk” AI systems making consequential decisions, including those in recruitment, hiring, and promotions that implicate protected characteristics. Early alignment with these standards can provide an opportunity to gain an edge in the talent wars.

Central to the proposal are requirements for independent third-party audits, conducted annually by both developers and deployers, algorithmic fairness across demographics, and remediation protocols. Public summaries would be mandated, subject to confidentiality for proprietary elements.

Enforcement through the attorney general carries $10,000 civil penalties per violation, alongside a private right of action for affected individuals. This extends the framework of New York City’s Local Law 144, which requires bias audits and candidate notifications for automated employment tools and has revealed inconsistent compliance in studies of hundreds of users.

While the bill addresses broader sectors such as housing and finance, its employment provisions resonate in a state where AI permeates tech-driven hiring. Business groups have raised concerns over audit costs, but proponents emphasize the rising tide of claims, with successful disparate impact suits underscoring the need for attention.

In addition, an amendment set requirements for ongoing risk management programs, increasing deployer responsibilities. Should the bill pass, we anticipate heightened scrutiny, potentially mirroring a projected uptick in litigation as AI tools expand.

Looking Ahead

Viewed together, these initiatives from New Jersey and New York reflect broader, state-driven accountability in AI governance, while emphasizing transparency and due diligence.

The challenge to employers lies in ensuring compliance with the demands of new operational strategies and requirements. The National Conference of State Legislatures reports active AI bills in every state this year, suggesting potential convergence but also interim complexity.

New Jersey’s enforcement-ready guidance and New York’s legislative blueprint offer employers a roadmap to responsibly harness AI’s potential.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Luis A. Ayala is an associate at Gordon Rees Scully Mansukhani and focuses his practice on employment law, providing strategic counsel and litigation defense to employers facing a broad range of workplace issues.

Write for Us: Author Guidelines