‹ Back

DoorDash Launches AI Training Tasks App for Couriers

 

DoorDash introduces a “tasks” app that lets couriers earn money by submitting videos and audio to train AI systems.

 

 

DoorDash has introduced a new standalone “Tasks” app that pays delivery couriers to complete video and audio assignments designed to train artificial intelligence and robotics systems.

 

The company said the app enables its network of delivery workers to perform short, structured activities—such as filming everyday actions or recording speech in different languages—in exchange for compensation determined by task complexity and effort.

 

Tasks App Expands Gig Work Beyond Deliveries

 

The launch marks a functional expansion of DoorDash’s gig economy model, allowing couriers to earn income outside traditional delivery work. In addition to the standalone app, similar assignments are being integrated into the existing Dasher platform, where workers can complete “digital tasks” between deliveries or on their own schedule.

 

Examples of these assignments include capturing photos of restaurant dishes to improve menu listings or documenting physical locations—such as hotel entrances—to assist navigation for future deliveries.

 

DoorDash stated in a company blog post that the initiative is intended to “help more businesses understand what’s happening on the ground and gather new insights,” while offering couriers “a new way to earn on their own terms,” according to Ethan Beatty, general manager of DoorDash Tasks.

 

The program leverages the scale of DoorDash’s workforce, which the company said includes more than 8 million couriers across the United States capable of collecting geographically diverse, real-world data.

 

Data Collection Targets AI and Robotics Training

 

DoorDash indicated that the audio and video data generated through the Tasks app will be used to evaluate and improve both its internal AI models and those of external partners across industries including retail, insurance, hospitality, and technology.

 

The assignments are designed to capture real-world scenarios that machine learning systems require to better interpret physical environments. One cited example involves a courier recording themselves washing dishes while wearing a body camera, ensuring each object remains visible for a defined period.

 

The company said such datasets help AI systems “understand the physical world,” highlighting the increasing reliance on human-generated, context-rich inputs to train advanced models.

 

Availability and Market Scope

 

The Tasks app and its in-app equivalents are currently available in select U.S. markets, with notable exclusions including California, New York City, Seattle, and Colorado. DoorDash said it plans to expand both the range of available tasks and geographic coverage over time.

 

Compensation for each assignment is displayed upfront, with payments calibrated to the effort and complexity required, according to the company.

 

Industry Trend Toward Workforce-Based AI Training

 

DoorDash’s move aligns with a broader shift among technology platforms seeking to use distributed workforces to generate training data for AI systems. Uber has introduced similar initiatives enabling drivers to complete microtasks such as submitting photos or recordings to support AI development.

 

The approach reflects increasing demand for large-scale, real-world datasets to support machine learning and robotics applications, with companies leveraging existing gig networks as a scalable data collection mechanism.

 

DoorDash’s introduction of the Tasks app formalizes this model within its platform, integrating data generation into its broader ecosystem of on-demand labor while extending the role of couriers beyond logistics into AI training workflows.