DoorDash has launched a new “tasks” app designed to pay its network of couriers to capture and submit real-world video footage, a move that signals the company’s deeper push into artificial intelligence infrastructure and data collection. The app allows drivers to complete short assignments—such as filming storefronts, menus, or parking conditions—providing valuable training data for machine learning systems that underpin logistics, mapping, and automation. While framed as an extension of gig work flexibility, the initiative highlights how major tech platforms are increasingly leveraging distributed labor to fuel AI development, raising broader questions about compensation, data ownership, and the long-term role of human workers in training systems that may ultimately reduce reliance on them.
Sources
https://techcrunch.com/2026/03/19/doordash-launches-a-new-tasks-app-that-pays-couriers-to-submit-videos-to-train-ai/
https://www.reuters.com/technology/doordash-expands-ai-data-collection-gig-workers-2026-03-19/
https://www.theverge.com/2026/3/19/doordash-ai-tasks-app-gig-workers-video-training-data
Key Takeaways
- DoorDash is monetizing its existing gig workforce by turning couriers into on-demand data collectors for AI training.
- The model reflects a broader industry trend of outsourcing AI data gathering to low-cost, distributed labor pools.
- The approach raises concerns about fair compensation and whether workers are indirectly training systems that could replace parts of their own roles.
In-Depth
DoorDash’s latest move into artificial intelligence development is not subtle—it is strategic, calculated, and very much in line with where the tech economy is heading. By launching a dedicated tasks app that pays couriers to capture video data, the company is effectively transforming its gig workforce into a decentralized data acquisition engine. This is not just about food delivery anymore; it is about building the infrastructure that powers future automation.
The concept is simple on the surface. Couriers receive assignments that might involve filming a restaurant entrance, documenting signage, or capturing environmental conditions like traffic flow or parking accessibility. But beneath that simplicity is a sophisticated objective: feeding machine learning systems with high-quality, real-world data that cannot be easily scraped from the internet. This kind of ground-level intelligence is essential for refining logistics algorithms, improving mapping accuracy, and ultimately enabling more autonomous operations.
What stands out here is how efficiently DoorDash is leveraging assets it already controls. Instead of hiring specialized data collection teams, it is tapping into an existing workforce that is already mobile, geographically distributed, and accustomed to task-based pay. From a business standpoint, it is hard to argue with the efficiency. From a worker standpoint, however, the picture is more complicated.
Gig workers are being offered another way to earn income, which on its face aligns with the flexibility many of them value. But the compensation structure and long-term implications deserve scrutiny. These workers are not just completing tasks; they are contributing to the development of systems that could eventually reduce the need for human involvement in delivery logistics. In other words, they may be helping build the very tools that could one day displace them.
This model also reflects a broader shift across the tech industry. Companies are increasingly turning to distributed labor to gather the massive datasets required for AI training. It is a cost-effective approach, but it raises legitimate questions about whether workers are being fairly compensated for the value they create. Data is the new currency in the AI economy, and those generating it are often the least empowered participants in the system.
At the same time, there is a pragmatic argument to be made. Innovation in logistics and automation is not slowing down, and companies that fail to invest in AI risk falling behind. DoorDash is positioning itself to stay competitive, and this initiative gives it a scalable way to accelerate development. The tension lies in balancing that drive for innovation with a fair and transparent relationship with the workforce that makes it possible.
Ultimately, this is a glimpse into the future of work—one where the lines between labor and data generation blur, and where everyday tasks double as inputs for increasingly intelligent systems. Whether that future benefits workers as much as it benefits companies remains an open question.