DoorDash Expands Beyond Delivery With Paid 'Tasks' That Require Workers To Record Daily Life For Training Humanoid Robots

DoorDash has quietly repositioned itself as a data company, launching a paid programme that turns its eight million delivery couriers into on-demand video contributors for the artificial intelligence and robotics industries.

On 19 March 2026, the San Francisco-based food delivery giant formally announced DoorDash Tasks, a standalone app and in-app feature set that pays couriers, known as Dashers, to complete short physical and digital assignments.

The announcement, first reported by Bloomberg, positions DoorDash as the latest, and arguably best-resourced, entrant in a fast-expanding market for human-generated, real-world training data at a moment when the robotics industry is desperate for exactly that.

How the Tasks App Works — And What Dashers Are Being Asked to Film

The Tasks product operates on two levels. The first is a new category within the existing Dasher app, covering what DoorDash describes as digital business intelligence: photographing restaurant dishes for menu listings, capturing hotel entrance points to assist future delivery accuracy, or scanning supermarket shelves to check inventory. The second, and more striking level is the new standalone Tasks app, which lists assignments involving the recording of domestic life.

The official DoorDash announcement confirmed that workers can film themselves completing everyday chores, or record conversations in languages other than English. Bloomberg reported specific task instructions that give a clearer picture of what the footage is actually for: one assignment asks couriers to wear a body camera aimed at their hands while scrubbing at least five dishes, holding each clean dish steady in frame for a few seconds before moving on.

Pay is displayed upfront per task and adjusted for complexity; harder assignments, such as pruning and repotting plants, earn more than simpler ones. No average rates or earnings floor figures have been published by the company. DoorDash has also folded its pre-existing Waymo partnership into the Tasks platform.

Confirmed by both DoorDash and Waymo to TechCrunch and Bloomberg in February 2026, that programme pays nearby Dashers approximately $14 (£11) to drive to a Waymo robotaxi and close a door left ajar by a passenger, a safety issue that prevents the autonomous vehicle from moving.

Ethan Beatty, General Manager of DoorDash Tasks, said in a statement published on the company's newsroom: 'It's simple: you can't deliver to a door you can't find or get someone milk if you don't know what's on the shelf. These are the kinds of real-world problems we've been solving for over a decade, and we realised the same capabilities that helped us could help other businesses too.'

DoorDash Enters the AI Data Market — and Arrives With Scale That Rivals Cannot Match

The launch places DoorDash in direct competition with dedicated data labelling and collection companies, including Scale AI, which built a substantial business on remote crowdsourced annotation workforces. The difference is distribution.

DoorDash's network of eight million couriers already operates across nearly every postcode in the United States, and the company has spent over a decade building the logistics infrastructure: dispatch, task verification, payment, that underpins the Tasks model.

The kind of data DoorDash is now positioned to supply is also increasingly scarce. As robotics developers push humanoid machines into domestic settings, they require footage of real people performing real household tasks in real environments. Simulated data, however sophisticated, does not fully replicate the variability of human movement and domestic geography.

NBC News reported that California-based Sunday Robotics distributes a 'skill capture glove' to volunteers nationwide, collecting motion data from household chores to train its home robot. Instawork, a staffing app, has separately been recruiting workers in Los Angeles to clean their homes whilst wearing a headband-mounted phone camera, per a report in the Los Angeles Times.

DoorDash is not the first major platform to recognise its gig workforce as a data asset. Uber announced a comparable initiative late last year, allowing US drivers to earn additional income by uploading photos used to train AI systems. Instacart has introduced similar programmes.

Privacy Questions and the Conspicuous Absence of Regulated Markets

The Tasks app is currently available only in select US markets. Notably absent are California, New York City, Seattle and Colorado, jurisdictions that carry significantly stricter data privacy protections and gig worker regulations than much of the rest of the country. DoorDash has not offered a public explanation for the exclusions, but the pattern is consistent with the company navigating regulatory exposure.

As The Next Web noted in its coverage of the launch, DoorDash has not published detail on consent practices, data retention policies, or the rights that workers retain over footage recorded in their own homes. For a programme that asks couriers to bring cameras into their kitchens and capture their own voices, those are not minor omissions.

The footage is intimate by nature; it depicts domestic environments, personal routines, the interiors of private residences — and the end use, feeding humanoid robot training models owned by unnamed third-party partners, extends well beyond what a courier agreement has historically implied.

The eWeek analysis of the launch observed a structural irony at the heart of the model: gig workers are among the groups most frequently identified as vulnerable to displacement by automation, and here they are being paid to accelerate the very systems that could reduce demand for human labour in domestic and logistics environments.

DoorDash has not addressed that question publicly, and Waymo's own position, that future vehicles will include automated door closures, at which point the door-closing gig disappears, makes the dynamic concrete rather than theoretical.

The exclusion of heavily regulated markets, the absence of data governance disclosures, and the per-task rather than hourly pay structure leave significant questions unanswered at launch — questions that regulators in excluded jurisdictions may eventually bring to bear on the programme as it expands.

DoorDash has built a business on the movement of food; it is now building one on the movement of data, and the distinction between the two is growing harder for regulators, workers and the public to ignore.