AI is accelerating the "gig-ification" of work.

Shen Translation is a translation team under 36Kr, focusing on technology, business, workplace, life and other fields, with a focus on introducing new foreign technologies, new perspectives and new trends.

Editor's Note: Artificial intelligence is reshaping the global labor market. Enterprises hope to improve efficiency and reduce costs with the help of AI, but at the same time, more and more jobs may be split into short-term, temporary and outsourced work. The gig economy once affected drivers and delivery workers, and now it is spreading to knowledge-based positions such as customer service staff, programmers and financial analysts. AI brings not only a productivity revolution, but also profound changes in labor protection and employment structure. This article is translated.

AI is Driving the Gig-ification of Work

In 2024, the buy-now-pay-later firm Klarna announced that it would cut hundreds of customer service positions and start using AI chatbots to replace human customer service representatives. This move is expected to save the company millions of dollars in costs.

But a year later, amid constant customer complaints about declining service quality, Klarna began rehiring human customer service staff. At first glance, this seems to be a victory for human workers in the AI era. However, the truth is far more complex than it appears.

Klarna did not rehire traditional full-time customer service employees. Instead, it signed a group of customer service personnel through external agencies. Subsequently, the company adopted a Ubers-like employment model, which Sebastian Siemiatkowski, CEO of Klarna, called "Uber-style work arrangements".

Today, most basic customer issues are still handled by AI chatbots, while a growing number of gig customer service representatives are responsible for dealing with more complex problems. Siemiatkowski said in a podcast program in February: "Just like a person can go to drive for Uber at any time, they can also log in to the system and provide customer services for Klarna."

This case can be regarded as a microcosm of how artificial intelligence changes the way of labor. Although labor economists still disagree on how many jobs AI will replace, they have basically reached a consensus that AI will replace part of the tasks in most jobs.

Optimists believe that AI can undertake a large number of repetitive tasks that human employees used to complete, freeing people up to engage in higher-value work. But another, more pessimistic view is that as enterprises integrate AI technology more deeply, they may use it to reduce the number of full-time employees, and instead build a more decentralized, more flexible labor system similar to the gig economy.

"Gig work" refers to flexible, short-term or on-demand work. The concept originally came from the music industry, for example, when a band completes a "gig" (performance). Today, the term is more often used to describe workers on platforms such as Uber, DoorDash, and TaskRabbit.

This type of work gives workers a certain degree of autonomy, and they can choose when and how long to work. But at the same time, they lose most of the benefits that full-time employees usually enjoy, including paid leave, medical insurance, work injury compensation, overtime pay, and even the minimum wage guarantee.

Alexandrea Ravenelle, a sociologist at the University of North Carolina at Chapel Hill and author of Hustle and Gig: Struggling and Surviving in the Sharing Economy, said: "Sociologists who study labor often discuss the trend that work is changing from 'career' to 'job' and then to 'gig'. And AI makes this process much easier."

Over the past decade, the gig economy has been largely associated with ride-hailing drivers and delivery workers. But now, more and more industries are finding that if enterprises can outsource part of their work content with the help of technology, they can reduce labor costs, and no longer need to recruit a large number of full-time employees with welfare benefits, but turn to hiring more contract workers.

This change is most obviously impacting white-collar office workers. The reason is that enterprises are working hard to prove that the adoption of AI can bring efficiency improvements. Mary Gray, senior principal researcher at Microsoft Research and author of Ghost Work: How Silicon Valley Built a New Global Underclass, pointed out that there is no evidence that jobs will disappear completely, but a large amount of evidence shows that once enterprises can break away from full-time employment relationships, they will do so.

Gray believes that technology can drive this transformation, but the main reason why enterprises adopt this model is still to reduce costs. Ravenelle said: "We will see this change in every industry. I don't think any industry can completely avoid this trend."

A Large Number of Jobs May Be Gig-ified

Fifteen years ago, gig platforms such as Uber, DoorDash and Instacart attracted workers with beautiful promises of freedom and flexibility. These platforms offer on-demand work opportunities to anyone who has a car and some free time.

Workers were told that they could be their own bosses, arrange their own working hours, and sometimes earn higher incomes than other jobs. However, the reality turned out to be completely different from these promises.

Today, millions of platform workers around the world find themselves in a highly unstable employment situation. Their income is unstable, working hours are unpredictable, and there are almost no labor protections at all.

A recent report released by Human Rights Watch details this global consequence. The report points out that the gig economy is depriving workers of some of their most basic rights, including the minimum wage, work injury compensation, paid sick leave, and the right to independently control working hours. At the same time, enterprises have reaped record profits as a result.

Although platforms tell workers that they are "their own bosses", in reality, platforms exert huge influence on their work through algorithms, including assigning tasks, setting compensation standards, and evaluating work performance.

Some delivery workers in the Human Rights Watch report said that they had been in traffic accidents, but because they could not get work injury compensation, they had to pay for their medical expenses themselves. For full-time employees in the United States, work injury compensation is usually a basic protection after they suffer occupational injuries. Other workers described a large amount of unpaid waiting time. They have to wait for a long time to pick up customers or receive orders, and it is estimated that even half of their working hours may not be paid.

Lena Simet, senior advisor on economic justice at Human Rights Watch and co-author of the report, said these phenomena should serve as a warning to all workers. She said: "What we see in gig work is, to some extent, the first sign of a broader trend."

"A lot of jobs are likely to be 'gig-ified'. I fully believe that this foreshadows possible changes in many areas of the future labor market." Some data show that this future has already arrived. A recent survey by the freelance platform Upwork shows that about 60 million Americans, accounting for 39% of the total labor force, are already engaged in freelance or gig work, including both full-time and part-time forms.

According to the forecast of the global data analysis platform Statista, this number may grow to 86 million by 2027, accounting for about half of the U.S. labor force. The fastest-growing and largest group is not ride-hailing drivers or delivery workers, but knowledge workers. They include customer service representatives, copywriters, financial analysts, paralegals, journalists, and programmers.

Once workers are defined as "contractors" rather than formal employees, the labor protections won by generations in the past may gradually disappear. Ravenelle said: "This means that the labor protections that generations of people fought for through arduous struggles are regressing. These protections, which our great-grandparents even paid their lives for, may all disappear now."

When Gig Work Becomes the Only Option

More and more workers are entering this work model, and some of the changes are driven by the adoption of AI technology by enterprises. Last month, Ravenelle published a study focusing on the impact of AI on workers in the creative industry. The research subjects include actors, writers, photographers, dancers, musicians, producers, costume designers and other groups.

Some creative workers believe that they will not be impacted by the AI wave, because they believe their work is too complex for large language models to complete, or their work relies on cultural judgment. But others have been forced into a "mercenary labor" state, which even includes work to help train AI systems.

One musician in Ravenelle's study chose one of the highest-paying jobs in the current market and became an "algorithmic composer". She is responsible for producing basic music clips and rhythms to train AI systems that may replace human musicians in the future.

Another subject of the study is a writer. He accepted a short-term project from a large technology company to evaluate the quality of AI-generated text. When asked if he was worried that he was training software that might replace himself in the future, the writer replied: "This is my best option. If you can think of a better one, please tell me."

Another actor in the study described his experience of participating in a project of a streaming entertainment giant. He was told that this work might eventually reduce or even eliminate the demand for extras and background actors in film and television works. Faced with the task of participating in a job that may directly reduce the employment opportunities of others, the actor had a strong moral conflict.

He said: "I feel like a construction worker laying bricks for a gas chamber." But since there were no other options, he could only accept the reality: "AI is taking over the world, and I am penniless. I can only go with the flow."

Workers enter the gig economy often because in an uncertain job market, it has become the best or highest-paying option they can find. In fact, this was the original promise of the gig economy. When no other company is willing to hire, you can at least drive for Uber. Now, in the labor market impacted by AI, people are turning to gig jobs again.

Mercor, an AI data training startup, hired more than 30,000 contract workers in 2025 to perform training tasks for the world's largest AI companies. These contractors include a large number of highly educated professionals, such as doctors, lawyers and bankers. They can get high hourly pay by improving the capabilities of AI models.

Nurses are also facing an increasingly obvious "gig-ification" trend. Over the past few years, some hospital networks have outsourced part of their core labor needs to AI-powered labor platforms such as ShiftMed, CareRev and Clipboard Health. These platforms are called "Ubers for the nursing industry".

The promotion methods they adopt are very similar to the early platform economy. Download an app, enjoy flexible working hours, bid to choose open shifts, and be your own boss. But reality has proved that this is not the case. Nurses who work using these platforms say they often get lower wages, have to compete with other nurses for job opportunities, and also have to carry equipment that should originally be provided by employers, such as stethoscopes and thermometers, on their own.

However, a nurse quoted in the relevant report said: "This is my best option. I have no choice." Similar to Uber, these nursing platforms have been seeking regulatory exemptions in multiple states, and have achieved some success.

At present, at least 17 states have proposed relevant legislation, hoping to recognize gig nursing platforms as "healthcare worker platforms" rather than traditional nursing institutions. This means they can be exempted from many regulatory requirements that protect workers, resulting in lower incomes for workers and reduced labor protections. Many of the so-called "Ubers of the nursing industry" platforms have been valued at billions of dollars.

Looking for a Way Out

Faced with this change in economic structure, some workers have begun to take action. One of the important ways is to use one of the few bargaining tools available to workers — forming trade unions.

In March this year, medical workers in California went on strike, protesting against the use of artificial intelligence technology by Kaiser Permanente and expressing concerns about outsourcing part of their work to technical systems. In May, information technology staff at the University of California voted to form a union. One of the main reasons they put forward is the concern about the risk of layoffs brought by artificial intelligence, and the hope to have more say in how AI will be used in the future.

Max Belasco, a business systems analyst at the University of California, Los Angeles (UCLA), who participated in the union organization work, said: "The concern about artificial intelligence is one of the important reasons why we decided to organize."

He added that most IT staff at the University of California are not against artificial intelligence or other technologies. They hope that the school can apply these technologies in a more strategic way, rather than simply using AI as a "cost-cutting tool".

But if we want to truly build a more comprehensive labor protection system, the actions of workers alone are far from enough. Researchers believe that more comprehensive policy reforms need to be promoted at the state, federal and international levels.

Mary Gray, senior principal researcher at Microsoft Research, said that in the future, we can consider establishing a "basic welfare" system. That is, no matter what type of work a person does, they can get basic protections, such as universal medical insurance or universal basic income. Another policy direction is to formally establish protection mechanisms for independent contractors and gig workers.

At present, the International Labor Organization of the United Nations is discussing a global treaty. If the treaty is finally established, it may set international standards on wage levels and work safety for gig workers. Lena Simet, an advisor at Human Rights Watch, believes that this could be an important step to improve labor protection.

But she also reminds that now is the critical period to push for better policy protection, and if we continue to delay, it may be too late in the future. She believes that governments of all countries have been too cautious in implementing relevant measures in the past. Governments tend to accept the claims of enterprises that stricter labor protections will affect the operational capacity of enterprises.

But Simet pointed out: "Even if they have to comply with some regulatory requirements, this is still a very profitable business model. If a business model can only survive by exploiting workers, then perhaps it should not exist at all."

Translator: Xiaochuan