AI Frontier Institute

Regulates use of artificial intelligence-based systems for electronic monitoring regarding employment and public services.

View primary source →

New Jersey S4075 (introduced May 4, 2026; reported from the Senate Labor Committee Oct. 1, 2026 and referred to Senate Budget and Appropriations) has not been enacted. If enacted, it would regulate AI-based electronic monitoring tools (EMT), automated employment decision systems (AEDS), and automated benefit or service decision systems (ABSDS) used by employers and public entities. Section 2 would prohibit uses that violate collective bargaining agreements or labor law, harm worker health, safety, privacy or dignity, or monitor workers in bathrooms and other private areas or while off duty. It would also bar requiring monitoring software on personal devices, implanted devices, or disclosure of device or social media passwords. Further prohibitions cover surveillance of residences or personal vehicles, harmful productivity quotas, adverse action based solely on keystroke or idle-time tracking, use of protected-class or union-related data, and retaliation. Section 3 would require that an AEDS or EMT meet limited allowable purposes and data-minimization rules and pass a pre-deployment impact assessment by an independent auditor (or the Department of Labor and Workforce Development for public employees). Section 4 would require a department impact assessment for an ABSDS. Vendors must supply design, data, accuracy and job-loss estimate information, and vendors would pay the department's direct costs for department assessments. Reports go into a public registry. Section 5 covers data sharing limits, accuracy, access and correction rights, retention limits, security, 48-hour breach notice, and joint liability. Sections 6 and 7 would require 60-day plain-language notice to affected workers, service beneficiaries and unions, and a required written response to union compliance concerns before implementation. Section 8 would require 10-day notice of adverse decisions, access to data, and an appeal to a human reviewer. Section 9 would require meaningful human oversight and bar decisions based exclusively on automated outputs. Section 10 would protect employees who refuse harmful AI outputs. Sections 11-13 would make public employer automated services agreements a mandatory subject of negotiation, restrict when they may be entered into, and give displaced public employees severance and health coverage (at least one year, two years for those with five or more years of service), seniority, recall rights and oversight-role training. Sections 14, 18, 19 and 20 provide enforcement through unfair practice charges, criminal and administrative penalties, civil suits with fines and damages, and joint liability for vendors and deployers. Section 21 directs the commissioner to adopt regulations, and section 22 sets an effective date 18 months after enactment.

← Back to the tracker