DoorDash rolls out “Tasks” to pay couriers for AI training data

Gig workers as data suppliers: a new revenue stream

U.S. delivery giant DoorDash has begun testing a feature called “Tasks” that lets couriers earn extra cash by performing simple micro‑jobs in gaps between deliveries. It has been reported that tasks include photographing restaurant dishes and recording natural conversations in non‑English languages — inputs DoorDash says will be used to train its own AI and robotics models and to support partners across retail, insurance, hotels and tech.

What the app does — and why it matters

Reportedly DoorDash is running Tasks as a standalone app where every microtask shows payment up front; the fee varies by complexity. This is familiar territory for AI firms that have increasingly turned to paid micro‑tasking to collect labeled data rather than relying solely on scraped or copyrighted material. In an industry frequently embroiled in lawsuits over unauthorized use of copyrighted content, paid collection offers a clearer compensation path — but not everyone agrees it solves broader consent and privacy questions.

A global pattern with local implications

Why should Western readers care? Because this move mirrors trends already visible in China’s AI ecosystem, where startups and big tech alike are experimenting with compensated data‑collection models to fuel large language and vision models. It also sits at the intersection of labor, privacy and geopolitics: how platforms pay and protect couriers, how multilingual datasets are sourced, and how U.S. and Chinese firms compete for the same scarce training material under diverging regulatory regimes. Is this innovation or exploitation? The answer will depend on transparency, worker protections and enforcement.

Platform notice

It has been reported that the original material was posted by a user on a third‑party platform; the platform states it merely provides information storage services.