In today’s ever-evolving work environment, discussions around workplace policies frequently center on the concept of “algorithmic management.” It’s suggested that companies are tightening their grasp on employees through various digital “tactics.” These tactics encompass everything from monitoring keystrokes to tracking locations and even surveillance via webcams. This narrative is prevalent in academic journals, government reports, and popular media. Notably, this narrative has permeated federal regulations, including the U.S. Department of Labor’s current rules regarding independent contractors. Like other popular accounts, this rule assumes that algorithmic management is widespread and treats it as a form of “control.”
However, there’s a significant flaw in this perspective: algorithmic management is more of a buzzword than a concrete reality. It refers to conventional business practices such as performance monitoring, providing incentives, and tracking work. Critics have successfully rebranded these longstanding practices as something new and detrimental to American workers, consequently influencing policies at the Department of Labor over the past few years.
Fortunately, a shift may be on the horizon. The Department of Labor has recently proposed to eliminate the term from its regulations, opting instead for a more traditional understanding of workplace “control.” This revised definition acknowledges that incentives and monitoring are not inherently forms of control. In reality, they represent methods businesses use to manage suppliers when they lack direct authority, acting as indicators of absence rather than evidence of control, even when labeled as “algorithmic management.”
Algorithmic Management: A Vague Concept
The term algorithmic management has always been somewhat ambiguous. It originated from a 2015 academic study that explored the then-emerging rideshare platforms. The research focused on how these platforms utilized digital tools to coordinate a large pool of independent drivers. It highlighted three key tools: matching riders with drivers, customer-rating systems, and surge pricing. The findings suggested that these tools effectively balanced driver availability with rider demand while maintaining driver satisfaction. This portrayal painted algorithmic management in a positive light.
However, the term has since evolved into something more ominous. It is now associated with a wide range of technologies, from scheduling software to chatbots and AI, the latter of which has escalated fears surrounding its implications. Commentators like Noam Scheiber from the New York Times, Sam Levine of the Federal Trade Commission, and Veena Dubal from UC Irvine have depicted algorithmic management in alarming terms, suggesting that companies exploit it to manipulate workers into longer hours for less pay. Dubal has even likened it to a modern Jim Crow.
Such rhetoric has tangible impacts on public policy. In 2024, the Department of Labor enacted a regulation differentiating employees from independent contractors, a distinction critical under the Fair Labor Standards Act (FLSA), which provides minimum wage and overtime protections exclusively to employees. The regulation outlined that algorithmic management plays a role in defining “control” over work, specifically referencing “technological” monitoring, such as GPS tracking, as a form of control—despite lacking additional context. The regulation suggested that simply monitoring work through technology constituted control.
This interpretation marked a departure from historical definitions, which viewed control as involving active measures: businesses had to instruct or enforce consequences for workers. Yet the 2024 regulation redefined control to include mere observation, implying that businesses could exert control simply by monitoring work through technological means.
Old Logic in a New Age
This line of reasoning misrepresents the situation. Monitoring, incentives, and similar tactics should not be interpreted as signs of control; rather, they indicate its absence. Economically speaking, these methods address the “principal-agent problem.” This issue arises when a business hires a contractor, and their incentives diverge: the business seeks optimal service at minimal cost, while the contractor wants maximum compensation for minimal effort. To safeguard its interests, the business must implement measures to counteract subpar work. This would be straightforward with employees, as the business can instruct them directly. However, with contractors, the business often lacks comprehensive information, thus necessitating methods such as progress reporting (monitoring) or offering bonuses for high-quality outcomes (incentives).
This indirect influence has never been classified as “control” in terms of worker classification—and it shouldn’t be. Otherwise, it would complicate the classification of independent work. All principals inevitably monitor their agents to some extent, meaning that monitoring alone provides no insight into the nature of the relationship. For instance, when a business arranges for package delivery and pays for confirmation, this constitutes a form of monitoring, but nobody would argue that every logistics provider operates as an employee of every client.
Modern technology doesn’t alter this dynamic. Even with technological tools, businesses still encounter principal-agent issues. The most effective strategies for addressing these challenges remain monitoring, incentives, and similar indirect methods. While technological advancements may enhance efficiency, the foundational dynamics remain unchanged—even when cloaked in alarming terminologies like “algorithmic management.”
Understandably, the efficiency offered by algorithmic tools raises concerns. Employees often fear that if employers monitor keystrokes or access webcams, this information could lead to excessive micromanagement. However, the crux of the matter lies not in the collection of data but rather in its application. If employers leverage keystroke information to evaluate, reward, or penalize workers, it’s the usage that constitutes control. Legally, the question revolves around whether the principal exerts control over the work—not whether they are privy to the methods involved. This holds true regardless of whether data collection occurs through technology, delivery confirmations, or traditional observation.
Fortunately, the Department of Labor appears to be aligning with this understanding. In late February 2026, they introduced a new rule clarifying the distinction between contractors and employees. This updated rule omits any mention of algorithmic management or technological control. Instead, it reverts to foundational principles: if an individual controls their work, they are likely a contractor; if they do not, they are probably classified as an employee. This principle remains applicable even in the digital realm.
In conclusion, the concept of algorithmic management is often overstated and misconstrued. As the Department of Labor moves toward a more traditional understanding of workplace control, it reaffirms that genuine control involves direct authority over the work performed, rather than mere monitoring or technological surveillance. By fostering accurate definitions, we can better navigate the complexities of the modern workforce.
