The Department of Labor’s Wage and Hour Division (WHD) has released a Field Assistance Bulletin (FAB) guiding employers on the use of artificial intelligence (AI) in human resources activities. The FAB focuses on how the use of AI can cause concerns complying with the Fair Labor Standards Act (FLSA), the Family and Medical Leave Act (FMLA), and the Employee Polygraph Protection Act (EPPA). The guidance predominantly addresses AI systems that undertake traditional human resources functions—such as tracking work hours, evaluating worker performance, setting schedules, and assigning tasks—and highlights their impact on remote and hybrid work environments across diverse workplaces.

• AI and the FLSA
  • Tracking Work Time: The FAB warns that while AI tools can track productivity, employers must compensate employees for all hours worked, not just productive time. Employers could unlawfully underpay workers if AI systems misclassify compensable hours due to flawed activity analysis.
  • Monitoring Break Time: Transitioning from worker-created time entries to AI-predicted entries poses risks to employers. The FAB notes that a change from timekeeping systems populated by worker-created time entries to timekeeping systems that make AI-generated predictions and auto-populated time entries also represents a potential compliance challenge for employers. For example, if an employee typically takes a 30-minute break every day, an AI system could learn to automatically deduct that regular 30-minute break. However, if it deducts that break so on a day when the employee works through it, the result would be the employer unlawfully failing to pay that employee wages for their actual work time.
  • Waiting Time: Employers must accurately track all time spent waiting for tasks, regardless of the technology used. With automated scheduling and task assignment systems, issues regarding hours worked may arise when employees wait for their next task or schedule update. If an employee is not given sufficient personal time, is not fully relieved from duties, or must stay near their workstation without a specific time to report back, they are typically considered “engaged to wait.” According to the FLSA, such waiting periods count as hours worked. AI systems might struggle to account for time as there would be no obvious input for a system to monitor.
  • Location-Monitoring: Using GPS to automate clocking in and out based on location can lead to undercounting work hours if the system only records time spent at the primary worksite. The FAB cautions that if a GPS-based AI system records only the worker’s time at the worksite as compensable work hours when the worker is performing work away from the worksite, it may fail to account for travel time between worksites or hours worked at other locations. Such processes could result in minimum wage or overtime pay violations. Employers using such systems must exercise responsible human oversight to ensure that they are properly accounting for and compensating employee work hours.
  • Calculating Wages Owed: AI systems that automate wage calculations must comply with federal wage requirements, including proper overtime pay. AI systems can use automated algorithms to determine pay rates based on supply and demand, customer traffic, location, worker efficiency, and task type. Employers must accurately determine the “regular rate” for overtime calculations when employees are paid different rates within a single workweek. Employers must ensure AI accurately determines pay rates, primarily when employees work at varying rates within the same week.

• AI and the FMLA
  • Processing Leave Requests and Certifications: The FMLA provides eligible employees of covered employers with job-protected leave for qualifying family and medical reasons. It requires continuation of their group health benefits under the same conditions as if they had not taken leave. Similarly, under the FMLA, an employer may require an employee to submit a certification from a healthcare provider to support the need for FMLA leave. Where AI is used to process leave requests, track time off, and integrate absence calendars, the FAB cautions that potential compliance challenges may arise. AI systems for processing leave requests and tracking time off must operate under strict human supervision. Errors in determining FMLA eligibility or mishandling certification requirements can lead to widespread violations if not adequately managed.

• Automated Scheduling Systems and Nursing Employee Protections
  • The PUMP Act: Under the FLSA, amended by the Providing Urgent Maternal Protections for Nursing Mothers Act (PUMP Act), AI systems must accommodate reasonable break times for nursing employees. Automated scheduling or timekeeping systems that limit the length, frequency, or timing of a nursing employee’s breaks to pump would violate the FLSA’s reasonable break time requirement.

• AI and the Employee Polygraph Protection Act (EPPA)
  • Lie-Detector Software: AI technologies using biometric analysis to detect deception fall under the prohibitions of the EPPA. The EPPA generally prohibits private employers from using lie detector tests on employees or for pre-employment screening of individuals. Employers cannot use any lie detector test, including any device that utilizes or incorporates AI technology on employees or job applicants, except under limited statutory exemptions.

AI technologies pose significant risks to employers of violating federal labor laws without vigilant human oversight. Employers must ensure AI systems are used responsibly to adhere to prevent potentially systemic and costly violations.

[View source.]