DoorDash Tasks App: Couriers Now Get Paid to Train AI
DoorDash has quietly crossed into new territory. The company best known for delivering burritos and groceries is now paying its couriers to film videos, record audio, and complete physical assignments — all to help train artificial intelligence systems. This is not a side experiment. It is a standalone app called Tasks, and it signals a major shift in how gig platforms are thinking about their workforce.
 |
| Credit: Pavlo Gonchar/SOPA Images/LightRocket / Getty Images |
If you are a DoorDash courier wondering whether the Tasks app is worth your time, or a tech observer tracking how AI companies are gathering real-world data, this article breaks down everything you need to know.
What Is the DoorDash Tasks App and How Does It Work
DoorDash announced the Tasks app on Thursday, March 19, 2026. The app is separate from the main DoorDash delivery platform and is designed specifically to let couriers earn money by completing short, structured assignments rather than food or package deliveries.
The assignments are designed to generate training data for AI and robotic systems. According to DoorDash, the goal is to help these systems better understand the physical world — things like how human hands move, what everyday environments look like, and how people speak in different languages.
Pay is shown upfront before a courier accepts any task, and the amount is based on the effort and complexity involved. This transparency is intentional. DoorDash wants workers to make an informed decision before committing to any assignment.
The app is currently available to existing delivery couriers, making it easy for DoorDash to onboard a large and geographically distributed workforce quickly.
What Kind of Tasks Are Couriers Being Asked to Complete
The assignments in the Tasks app are more hands-on than you might expect. One specific example involves a courier wearing a body camera and filming themselves washing at least five dishes, holding each clean dish in front of the camera for a few seconds before moving to the next.
Other tasks include recording audio in different languages, capturing footage of everyday physical actions, and completing activities that help AI systems map and interpret real-world movement. These are not abstract digital tasks. They require physical participation in a controlled but natural setting.
The variety of tasks reflects the diverse needs of DoorDash's AI partners. The footage and audio collected will be used to evaluate both in-house AI models and systems developed by partners operating in retail, insurance, hospitality, and technology sectors.
This means the data being gathered has commercial value far beyond food delivery. Couriers are essentially becoming field researchers for an expanding network of AI development projects.
Why DoorDash Is Turning Its Delivery Fleet Into an AI Training Network
The move makes a lot of strategic sense when you look at what AI developers actually need right now. Building capable AI systems — especially those that interact with the physical world — requires enormous amounts of high-quality, real-world data. That data is expensive and time-consuming to collect.
DoorDash already has tens of thousands of couriers spread across cities, suburbs, and neighborhoods of all kinds. These workers understand how to follow structured instructions, operate under time constraints, and navigate physical environments. They are, in many ways, an ideal field data collection network.
By launching the Tasks app, DoorDash turns an existing labor resource into a new revenue stream while also offering its workforce an additional income opportunity. It is a clever alignment of interests — at least on the surface.
The company is also positioning itself as a data and logistics infrastructure partner, not just a delivery middleman. That is a significant pivot in identity and business model.
DoorDash Is Not Alone — Gig Platforms Are Joining the AI Data Race
DoorDash is not the first gig economy company to explore this territory. Toward the end of 2025, a competitor platform announced plans to let its drivers earn supplemental income by completing small AI-related jobs, such as uploading photos to help train image recognition models.
The pattern is becoming clear. Platforms with large, flexible workforces are recognizing that their biggest asset is not just logistics capacity — it is human presence and physical mobility at scale. AI companies need data from the real world, and gig workers can provide it in ways that automated systems currently cannot.
This convergence of gig labor and AI development is likely to accelerate. As more AI systems move from text-based applications into physical robotics and real-world navigation, the demand for grounded, embodied training data will only grow.
For workers, this creates a genuinely new kind of opportunity. But it also raises questions about how their data, likeness, and labor are being used and compensated over the long term.
What This Means for Gig Workers and the Future of AI Labor
For couriers who are already part of the DoorDash ecosystem, the Tasks app offers something valuable: flexibility and upfront pay clarity. Unlike delivery shifts where earnings depend on tips, traffic, and order volume, Tasks assignments show the compensation before you commit. That is a meaningful improvement in pay transparency.
However, the broader implications deserve careful thought. When a courier films themselves washing dishes while wearing a body camera, that footage does not just disappear after training. It becomes part of a dataset that can be used, licensed, and potentially monetized across industries. The courier receives a one-time payment. The data may have a much longer and more lucrative life.
There is also the question of what this trend means for AI development ethics more broadly. Sourcing training data from a gig workforce creates efficiency, but it also means the people doing the work often have limited insight into how their contributions will ultimately be used or what systems they are helping to build.
For now, participation in the Tasks app appears voluntary and well-compensated relative to the effort involved. But as these programs expand, workers, advocates, and regulators will likely push for clearer disclosures and stronger protections.
AI Needs the Physical World, and Gig Workers Have It
One of the most fascinating aspects of the DoorDash Tasks launch is what it reveals about the current limits of artificial intelligence. Despite extraordinary progress in language models and image recognition, AI still struggles to understand the physical world with the nuance and reliability that humans take for granted.
Teaching a robot or an autonomous system to recognize a clean dish, understand spoken language across accents, or navigate a kitchen confidently requires data that no synthetic dataset can fully replicate. It requires real people, in real spaces, doing real things.
DoorDash couriers, by the nature of their work, already move through the physical world constantly. They navigate apartment hallways, interact with strangers, operate in all kinds of weather, and adapt to unpredictable environments. That lived, embodied experience is exactly what modern AI development desperately needs.
The Tasks app is essentially a bridge between the messiness of the physical world and the structured requirements of machine learning. And it is being built on the backs — and hands, and voices — of gig workers who may not fully realize how central they are to the next chapter of AI development.
What Comes Next for DoorDash Tasks and AI Data Collection
It is reasonable to expect the Tasks app to expand significantly in scope and scale. As DoorDash's AI partners identify new data needs, the range of assignments will likely grow. Tasks could evolve to include environmental scanning, object labeling, language translation, and far more complex physical demonstrations.
DoorDash may also look to open the platform beyond its existing courier network, potentially recruiting participants specifically for AI data collection rather than delivery work. That would represent a further evolution away from the traditional gig model and toward something more like a distributed research platform.
The intersection of gig work and AI development is one of the most consequential labor trends of this decade. DoorDash Tasks is an early and visible example of what that intersection looks like in practice. It will not be the last.
Whether this shift ultimately benefits workers, reshapes the economics of AI development, or raises new concerns about data rights and labor practices, one thing is certain: the line between delivery courier and AI data contributor has never been thinner.
