How AI and Automation Are Quietly Erroding the Global Gig Economy Safety Net
The gig economy, once seen as the future of work, faces a reckoning as AI automates white-collar freelance tasks and physical automation threatens delivery and driving roles.
Both supporters and critics of the gig economy have often presented it as the future of work. But has it actually placed workers on a fast track to being replaced by automation? This year brought a moment of reckoning for freelance platforms in white-collar fields, such as Fiverr or Upwork, which connect companies with independent contractors worldwide for well-defined projects like copywriting or graphic design.
It turns out that the very fragmentation of jobs into tasks that can be performed remotely, without the need for internal organizational knowledge or personal relationships, has made this type of work particularly vulnerable to AI replacement. Fiverr reported in the second quarter of this year a "clear slowdown in user traffic and overall demand in the marketplace," according to prepared remarks by CEO Micha Kaufman. He attributed this, among other things, to "recent model updates across a variety of large language models," stating simply: "AI automates simple tasks."
Upwork reported in the first quarter an "acceleration in AI adoption," which "impacted activity volume from lower-end market clients, affecting contracts of $500 and below." Although the heads of both platforms are trying to shift toward higher-level work that is less vulnerable to automation, investors remain unconvinced. Shares of Fiverr and Upwork have fallen by 61% and 47%, respectively, over the past year.
The Gig Economy as a Safety Net for Blue-Collar Workers
Meanwhile, physical automation threatens a series of roles previously filled by platform workers, including driving and package delivery. In June, the head of one of China's largest e-commerce groups warned that 700,000 of its delivery workers would be replaced by robots "sooner or later." In the US and China, drivers working on platforms like Uber and DiDi view the arrival of autonomous vehicles with deep apprehension. Uber has even begun acting as a lobbyist for drivers—an irony not lost on the drivers themselves—pushing for a slower rollout of autonomous cars to soften the transition.
The company also wants to promote rules that would require ride-hailing platforms to operate "hybrid networks" combining autonomous cars with human drivers. This is, of course, a self-serving move by Uber, which is trying to protect its business model from fully autonomous platforms. But this does not mean it is wrong when warning that robotaxis could lead "in the long run to less work for many drivers, including people who rely on platform work as a flexible safety net."
The Impact on the Labor Market and Future Trends
Gig platforms have indeed become a safety net for people who have been pushed out of the formal labor market or struggled to enter it in the first place. A 2023 World Bank report estimated that worldwide there are between 154 million and 435 million online platform workers—between 4.4% and 12.5% of the global labor workforce. Among the benefits of this phenomenon, according to the report's authors, is that it "helps cope with income shocks" and "serves as a kind of unemployment insurance."
China is perhaps the best example of a country where platform work has served as a labor market shock absorber. Amid the ongoing crisis in the construction sector and the growing automation of manufacturing jobs, the number of people working in China as food couriers or ride-sharing drivers grew by 10 million within two years, reaching 53 million in 2025, according to estimates by the China New Employment Forms Research Center, a Beijing-based research institute.
The impact is expected to happen beneath the radar: fewer tasks and lower incomes, rather than job losses of the kind economists are used to measuring and tracking. An illustration of this is the scarcity of data on the impact of robotaxis on drivers. However, a study by Gridwise, a driver assistance app for ride-hailing platforms, indicates that earnings per trip for drivers in US cities where autonomous vehicles operate declined between July 2024 and July 2025 relative to the national average.
Upwork provides a glimpse into what such platforms might look like when supply and demand are out of balance. Workers on Upwork are required to pay to submit job bids through a system called Connects. They can also pay more to promote their proposal to the top of the bid list. The threats to these specific workers and platforms do not necessarily herald the end of the business model underlying the gig economy.
New platforms are already beginning to emerge that appeal to other markets—whether in fields that are harder to automate, such as healthcare, or in white-collar fields where artificial intelligence itself creates new demands. One example is Mercor, which pays lawyers, journalists, and bankers, among others, to help improve and train AI models to do their work—thereby perhaps hastening their own replacement. Yet new platforms and new types of work will not necessarily employ the same people being pushed out of the old platforms.
The gig economy was never a particularly good safety net because it leaves people without the rights and protections of salaried employees, and with only partial state support regarding pensions and allowances. But now a new threat has emerged: robots are quietly chewing holes through this safety net. Policymakers may not notice until the ground gives way entirely beneath them.
