Computers & AI-powered algorithms
AI and Workplace Monitoring
All Along the New Watchtower: Artificial Intelligence, Workplace Monitoring, Automation, and the National Labor Relations Act
–Bradford J. Kelley, Littler
Workplace Technologies: A Brief Overview
Many workplace technologies use AI, which can be best understood as computer systems and algorithms “to perform tasks that typically require human-level intelligence to optimize aspects of the workplace, including enhancing productivity, streamlining operations, and improving decision making.” In most situations, employers engage third-party vendors that develop and sell AI-powered algorithms to perform a wide variety of human resources tasks. As a general rule, employers—not vendors—are generally liable for unlawful employment decisions. The NLRA is no exception to this rule, as it likewise tends to hold employers liable even where the technologies being sold by vendors might be responsible for the violation.
Workplace Technologies and the Employment Lifecycle
Workplace technologies are being used at all stages of the employment lifecycle. Research has consistently shown that AI tools properly used for employment decision-making may result in greater diversity, unbiased promotion decisions, and better retention of employees. Automated employment decision tools include algorithms that analyze social media information to determine which candidates see a job advertisement, tools that analyze the words in resumes, and programs that assess a candidate’s personality traits.
AI-enabled technologies are adding new ways to make a workplace more directly accessible to people with disabilities. Notably, “wearable technologies . . . have been [shown] to mitigate the effects of certain disabilities, thereby broadening employment opportunities for disabled workers while simultaneously preventing work-related accidents and improving productivity by reducing absences due to disability and illness.”
On a related note, AI is also helping disabled workers with reasonable accommodations in the workplace.
AI programs are increasingly used to identify, monitor, and measure employee performance. Historically, employers could more easily monitor their employees’ attendance and performance when they were all in physical locations on a daily basis, but many employers have turned to monitoring software to balance the proliferation of remote work.
Indeed, studies show that around 80% of large employers are using some type of monitoring software at the workplace. AI and machine learning are frequently used to track worker on-site and remote activities, including employee log-in times, idle time, overall computer usage, documents accessed, online activities, and are being used to measure employee performance. AI tools can also monitor whether employees are paying attention to their computer screens using webcams and eye-tracking software; employers also regularly monitor workers’ activities by installing spyware and GPS trackers on desktops and company-issued laptops.
These systems allow for companies to monitor the speed and location of drivers, including truck drivers and ride-sharing drivers working for platform-based systems.
These systems can also be used to verify if such workers gather in particular locations. Employees failing to achieve set performance standards might face formal disciplinary actions, including potential job termination, as suggested by algorithms.
Numerous employees, whether working from home or in other settings, are subject to tracking tools, scoring systems, so-called idle buttons, in addition to other types of ways to accumulate data. A pause in activity (whether real or perceived) can lead to penalties, including lost pay and termination.
In a similar way, AI is increasingly being used to measure or enhance worker productivity and efficiency.
For instance, some large transportation companies have used wearable technologies including ring scanners to help workers with package sorting, pickup, and delivery.
AI is also being used to incentivize worker productivity. For instance, some platform-based services use AI to incentivize driver productivity based on predictive analysis and tailored incentives. Retail stores use AI to evaluate and incentivize workers based on an automated analysis of their interactions with customers. Monitoring has spread among white-collar jobs and roles that usually require graduate degrees. Doctors, architects, academic administrators, nursing home workers, and others frequently describe that electronic surveillance is increasingly used to monitor their activities and behavior every minute of the day.
For instance, certain radiologists are shown scoreboards that display their periods of inactivity and how their productivity compares to their colleagues.
Software programs can generate timecards and calculate employees’ pay.
One company offers software to monitor remote workers by taking screenshots of their computers at set intervals and collect data, including keyboard activity and application use, to generate a timecard every ten minutes.
The timecard then creates a logbook for the workers and their managers that shows how the worker spent their time.
Another business reportedly uses software that generates a photo of employees’ faces as well as screenshots of the employees’ computer screens every ten minutes throughout the workday. The company then uses that information to pay the employees and other workers only for the time when the system detected them to be actively working (e.g., moving a mouse or a keystroke) based on the photos.
If the photo captures an employee during a brief moment of inactivity (e.g., a short coffee break of around thirty seconds or a quick bathroom break) such periods of perceived inactivity are considered non-compensable idle time so the system would dock an employee’s pay for the entire ten-minute duration.
The use of monitoring software to measure work time has already generated litigation. Most notably, in Kraemer v. Crossover Market, LLC, a plaintiff brought a putative collective action for unpaid off-the-clock work against her former employer and a company that operates a recruitment platform to hire and manage workers. The complaint alleged that the defendants required the plaintiff, an independent contractor who worked remotely, to install tracking software, and she was compensated based on the software’s tracking of her activities.
The plaintiff claimed that the system failed to account for various offline work, including reviewing and annotating hard-copy documents, receiving work-related phone calls away from her computer webcam, and participating in Zoom conferences on her mobile phone away from her workstation.
“[T]he complaint alleged that the ‘spyware software would not give credit for Plaintiff’s work when it did not detect her sitting in front of the computer, keystrokes on her keyboard or movement of her mouse.’” The parties settled for an undisclosed amount. Practitioners have explained that the lawsuit highlights a practical lesson for managing remote workers: “[T]here can be a problematic disconnect between what surveillance software is capable of measuring and the actual range of tasks an employee regularly performs in the course of carrying out their job duties, particularly where ‘offline’ activities are involved.”
READ MORE>> AI and Workplace Monitoring
About Bradford J. Kelley:
Brad is an internationally recognized workplace AI authority. He advises clients on how to maximize the benefits of using AI in the workplace while minimizing potential legal and business risks.
Brad has a broad practice representing employers in employment anti-discrimination and wage and hour matters. He focuses on advising clients about emerging technologies, including artificial intelligence (AI), and their impact in the workplace.