Indian factory workers train AI systems that may replace them
The factory floor is not a classroom, but the supervisors are acting as if it is - handing out head-mounted cameras to garment workers in India as if these devices were nothing more than a new piece of shared equipment, like a measuring tape or a cutting table. Yet what is being measured, recorded, and ultimately enclosed is not just a worker’s motion, but the very process by which a garment is made: a living commons of skill, rhythm, and tacit knowledge that has been built over generations. The commons was never a blank slate; it was the accumulated cooperation of hands that knew how to stretch fabric without puckering, how to align seams without shifting, how to finish a hem in half the time without sacrificing quality. Now the fence is being erected in real time: the head-mounted camera is the first post, and the AI model it feeds will be the gatekeeper. The question the supervisors do not ask - because they do not want to hear it - is who will stand inside that gate once the fence is complete.
Consider the scene: a worker in a Chennai factory, perhaps a woman in her thirties with calloused fingertips and a posture honed by years of leaning over sewing machines, is handed a device that will record every pivot, every hesitation, every micro-adjustment of her hands. The supervisors tell her it is for “training data,” as if the commons she has long stewarded were merely a raw material to be mined and refined into something proprietary. The irony is not subtle: the same hands that have, for decades, passed their skill down through apprenticeships are now being asked to gift it to an algorithm that may render those very hands obsolete. The factory management, of course, frames this as progress - consolidation, maturation, the natural arc of industry. But look closely: what is being consolidated is not capability, but control. The shared resource - this intricate, living choreography of making - is not being improved; it is being privatised. The fence is not a wall around a building; it is a wall around a process, with the gate locked to those who once moved freely within it.
The beneficiaries are already visible in the shadows. Who profits from the enclosure? Not the workers, who are being asked to fund their own redundancy. Not the Indian economy, which exports garments but imports dependency. The real beneficiaries are the owners of the AI models, the shareholders of the tech platforms, the venture capitalists who will monetise the data these cameras capture - selling back to the very factories the insights that once circulated freely among the workers themselves. The workers are being asked to treat their own labour as a sacrifice to the altar of efficiency, as if cooperation were a finite resource that must be surrendered to be counted. But cooperation does not deplete when it is shared; it multiplies. The workers’ collective knowledge is what made the factory run smoothly in the first place - yet now that same knowledge is being funnelled into a system designed to extract it, package it, and sell it back to the factories that once depended on it.
The supervisors will argue that the cameras are necessary for “safety” or “quality control,” the same tired pretext used for decades to justify enclosure in the name of oversight. But safety for whom? The factory management, already positioning itself to license the AI model to other factories, will claim the data is anonymised - yet anyone who has ever worked in a garment factory knows that no dataset can erase the embodied knowledge of a worker’s hands. The anonymisation is a fiction, a polite fiction that allows the enclosure to proceed under the guise of neutrality. The real safety being secured is not for the workers, but for the model’s owners, who can now claim that the commons they are enclosing was never really a commons at all - just a collection of raw materials waiting to be refined into a product.
What has been lost in this transaction is not just a job, but a way of knowing. A worker who has spent a decade perfecting the tension on a zipper does not merely lose a task when an AI model takes over; she loses the authority to define what counts as quality, what constitutes a flaw, what makes a garment worth making. The camera does not record her judgment; it records her compliance. The mutual aid that once flowed through the factory floor - older workers teaching newer ones, teams adjusting their rhythm to accommodate a slow stitcher - is being replaced by a system that treats knowledge as a commodity to be extracted and owned. The fence does not just keep people out; it rewrites the rules of who gets to decide what is valuable.
The garment workers of India are not being asked to wear cameras for their own good. They are being asked to wear shackles disguised as tools, to trade their collective wisdom for the promise of a future they will not recognise. The factory supervisors may call this consolidation, but it is nothing of the sort. It is enclosure. It is theft. And it is dressed up as progress because, as always, the fence is built by those who plan to stand on the other side.