In late April, the International Labour Organization published a working paper, but hardly anyone outside Geneva seemed to take notice. Depending on how skeptical you are about institutional timing, the fact that it arrived the day before International Workers’ Day felt either coincidental or intentional. The results were not nuanced. The study contended that AI-driven management is changing the psychosocial workplace in ways that preexisting safety frameworks were never intended to capture. Within 48 hours, the press cycle was over.
You can see why. There isn’t a single villain, no executive perp walk, and no leaked memo. Just a gradual build-up of minor intrusions, such as dashboards that track keystrokes, scheduling systems based on opaque logic, and hiring tools that reject applicants for reasons that are incomprehensible to human desk staff. This is the type of story that doesn’t become popular. Like office fluorescent lighting did decades ago, it simply fades into the background of working life and awaits normalization.
Working Paper 170, a rather somber title, is worth reading if only because it identifies the feelings that the majority of workers already have. The scale of surveillance has increased. Autonomy has decreased. A person’s day and occasionally their career are shaped by decisions made by systems that they are unable to question. Although the ILO’s framing is meticulous and bureaucratic, there is something almost concerning hidden beneath the policy language. According to the report, 27% of mid-level managers are concerned about the inadequate protection of employees’ physical and mental health. Something is wrong when the managers are uncomfortable.
You can currently see the texture of almost any mid-sized office in Dubai, Manila, or Mumbai. The screens are identical. The coffee makers are identical. However, a server in the rear of the building is recording the duration of each call, rating the tone of voice, and identifying slow responders. Last year, a warehouse employee at a Gulf logistics hub told a reporter that it was like being watched by an unblinking eye. I’ve never forgotten that phrase. observed by an unblinking entity.
The ILO recognizes a paradox, but it doesn’t fully address it. The same instruments that increase productivity levels are also undermining the conditions that initially enable quality work. Have faith. discretion. the silent satisfaction of being able to solve problems. High automation may actually increase mental workload while decreasing situational awareness, according to researchers cited in the paper. This may seem counterintuitive until you’ve seen a coworker stare at three dashboards in an attempt to determine why the algorithm has flagged her once more.
As expected, regulation is inconsistent. In their workplace guidelines, Singapore and Australia are beginning to address psychosocial risk. The EU’s AI Act is a step in the direction of worker protections, but it doesn’t quite corner them. As is typical in a growing market, adoption is moving quickly in India and the majority of the Gulf, while safeguards are moving more slowly. Depending on where your employer is based, the outcome could be a sort of geographic lottery for workplace dignity, the ILO cautions.
It’s difficult to ignore how this resembles previous transitions. The early twentieth-century factory floor. The late 20th-century open-plan office. The productivity story always came out years ahead of the health story. People inside the buildings had no choice but to adapt each time. Whether someone with regulatory authority determines that this paper is more valuable than a press release will determine whether the algorithmic workplace adheres to the same script. The silence has been louder than it ought to be thus far.
