New division leverages Uber’s expertise in gig work to tackle high-demand AI training tasks
Uber Technologies (UBER) has launched a new division called Scaled Solutions, expanding into the AI data labeling sector by leveraging its established workforce model to address growing demand for high-quality data for artificial intelligence training. Originally an in-house team handling data annotation for Uber’s core businesses—rideshare, food delivery, and freight—Scaled Solutions has now started offering its services to external clients, including notable partners like autonomous driving firm Aurora Innovation Inc. and augmented reality game developer Niantic Inc.
The AI data labeling sector is seeing explosive growth as companies worldwide require annotated datasets for machine learning models to make accurate predictions. Scaled Solutions enters a competitive space alongside major players like Scale AI, valued at $14 billion, which dominates data labeling for industries such as autonomous driving and natural language processing. Through this new venture, Uber aims to leverage its experience in recruiting gig workers to meet the needs of companies reliant on massive data sets for model training.
Uber’s move into data labeling is strategic, reflecting the company’s capacity to organize and manage large pools of contractors. With Scaled Solutions, Uber has begun onboarding skilled contractors across India, the United States, Canada, Poland, and Nicaragua. Contractors are paid per completed task, with payouts occurring monthly. Tasks range from labeling images and videos to evaluating complex AI outputs, with Uber tapping into skilled labor markets where remote work is highly valued.
In the U.S. and India, Uber’s Scaled Solutions team already handles various tasks critical to the rideshare and food delivery business. This includes validating maps, translating app content for local markets, and digitizing restaurant menus for Uber Eats. The team has also been instrumental in quality assurance, testing Uber’s apps across thousands of devices. These internal tasks laid the groundwork for Scaled Solutions to begin offering similar services to external clients like Aurora, where Uber’s contractors help classify objects for autonomous vehicles, and Niantic, where they aid in creating a 3D map for augmented reality games.
Uber’s entry into data labeling isn’t without challenges. The gig economy is under scrutiny for the compensation and working conditions of outsourced labor, especially for workers in developing countries. For instance, reports on similar platforms such as Remotasks reveal wage disparities, with some U.S.-based workers reportedly earning $18 an hour, while workers in Nairobi, Kenya, were paid around $10 for eight hours of annotation work. Scaled Solutions has not disclosed specific wage details across its international locations, and rates for tasks like evaluating AI-generated coding solutions can vary significantly.
To address the unique needs of AI model development, Uber has been hiring freelancers with specialized skills, such as programming or linguistic abilities, to provide feedback on AI-generated outputs. One India-based engineer shared that their tasks involved assessing AI solutions to coding problems, focusing on factors like functionality, efficiency, and code formatting. For completing three questions, the engineer received a payment equivalent to $2.37, illustrating the varied pay rates associated with these annotation tasks.
With Scaled Solutions, Uber aims to carve out a niche in a booming industry, backed by its expertise in flexible work opportunities and access to global labor pools. As AI becomes a cornerstone of technology development, Uber’s expansion into data labeling positions it as a key player in the ecosystem of AI model training, albeit within a complex landscape of labor ethics and evolving industry standards.
You might like this article:Gold Slips as Dollar Gains Amid Trump’s Tariff Threats
