Opinion: Responsible AI requires labor protections for ‘taskers’

Are you familiar with the term “tasker”?

They are the hundreds of thousands of people around the world performing a kind of piece-work in training and fine-tuning AI models. They label images, rank chatbot responses, write code, try to find gaps and flag inaccuracies.

Read more Letters:

Some of the work requires domain expertise; much of it doesn’t. Without their labor, none of the AI products or services that we are all using (whether intentionally or not) would exist.

Yet, while so many of us would recognize the names of major AI companies or their founders or CEOs, we do not often hear about taskers, and most people are much less familiar with the names of the companies that employ them, in a particular kind of contract employment.

Maybe that’s starting to change, a bit: Wired magazine, for example, recently published an insider account written by a tasker, detailing the stop-start-disappear nature of the projects, the unpaid hours of onboarding, the vague rubrics on the basis of which efforts are judged, the various kinds of unmet promises, and the “fever dream of aimless, directionless urgency” within which so many people of various ages and backgrounds are now spending their working hours (often as all-nighters).

But the article’s headline didn’t do it justice.

I wish it had been something more like “How Your Kids, Students, and Neighbors are Being Worn Down for Little Pay and No Security While AI Firms Benefit From Their Work.” Not catchy, I realize.

In a recent papal encyclical, Pope Leo XIV writes extensively about the need to preserve human dignity in the time of AI, and he specifically refers to taskers — though not by that name. “A significant part of the digital economy’s functioning relies on the silent work of millions of people engaged in essential yet largely unseen activities,” he notes, “such as data labeling, model training and content moderation, often involving disturbing material.” In a different section of Magnifica Humanitas, he adds, “It is not enough to invoke efficiency, nor to celebrate the benefits of innovation, if they are built on a chain of exploitation that remains deliberately hidden.”

Will the Pope’s highlighting of such labor make a difference? In 2024, Noema magazine published an article about the invisible work force. Before that, in 2023, an article called “AI Is a Lot of Work,” dove deeply into the same issues, detailing people whose work was “stripped of all its normal trappings: a schedule, colleagues, knowledge of what they were working on or whom they were working for. In fact, they rarely called it work at all — just ‘tasking.’” In 2019, researchers Mary Gray and Siddharth Suri co-authored a book titled Ghost Work, also focused on the “invisible human workforce.” Is now the time for this growing workforce to become visible?

A papal encyclical commands a very different type and scale of audience than any media coverage. And its calls to action will continue to resonate far longer.

But, also, times have changed, even from what they were just a couple of years ago. There are many more calls for regulation of various aspects of AI. There are lots of conversations about “AI governance.”

And there are now many, many programs, in universities and other organizations, focusing on “responsible AI.” Any such programs should address the role of taskers, and the need for protection of human dignity and the dignity of labor.

Does that sound too aspirational?

Surely not more aspirational than the search for AGI. As companies announce that they will pour hundreds of billions of dollars into building infrastructure, and communities are mobilizing to demand more in response, maybe the time has also come to make visible — and more properly protect and reward — the human infrastructure of AI.

Irina Raicu directs the Internet Ethics program at the Markkula Center for Applied Ethics at Santa Clara University. Her work addresses a wide variety of topics — from online privacy to social media’s impact on society, and the many ethical issues associated with artificial intelligence.