A viral Reddit post alleging exploitative practices at a major food-delivery company was later found to be AI-generated, misleading millions of readers, according to a report by TechCrunch.
The post, published by a Reddit user under the name Trowaway_whistleblow, claimed to come from a whistleblower working as a developer in the US-based food delivery platform. While the post initially avoided naming the company, it later implied the claims were connected to UberEats.
The user alleged the company operated a “rigged system” that exploited drivers through hidden algorithms. It also claimed that priority fees and driver benefits were never actually paid to drivers, and that the platform used a secret “desperation score” to track drivers under financial pressure and manipulate order assignments.
The Reddit post received 87,000 upvotes, and was shared across platforms such as X, securing 208,000 likes and 36.8 million impressions, the TechCrunch report stated.
However, the claims were later debunked by journalist Casey Newton of the tech publication Platformer. After initially engaging with the user, Newton concluded that the supposed evidence, including an 18-page “internal document” and an UberEats employee badge, was fabricated using generative AI.
Further analysis using Google Gemini confirmed that the images shared by the user carried SynthID watermarks, a tool designed to detect AI-generated or manipulated visuals even after cropping or compression.
The incident highlights growing challenges around trust and verification as artificial intelligence becomes more deeply embedded in the food delivery ecosystem.
In a separate case reported by TechCrunch, delivery platform DoorDash recently banned a rider for submitting an AI-generated image to fake a completed delivery.
The post, published by a Reddit user under the name Trowaway_whistleblow, claimed to come from a whistleblower working as a developer in the US-based food delivery platform. While the post initially avoided naming the company, it later implied the claims were connected to UberEats.
The user alleged the company operated a “rigged system” that exploited drivers through hidden algorithms. It also claimed that priority fees and driver benefits were never actually paid to drivers, and that the platform used a secret “desperation score” to track drivers under financial pressure and manipulate order assignments.
The Reddit post received 87,000 upvotes, and was shared across platforms such as X, securing 208,000 likes and 36.8 million impressions, the TechCrunch report stated.
However, the claims were later debunked by journalist Casey Newton of the tech publication Platformer. After initially engaging with the user, Newton concluded that the supposed evidence, including an 18-page “internal document” and an UberEats employee badge, was fabricated using generative AI.
Further analysis using Google Gemini confirmed that the images shared by the user carried SynthID watermarks, a tool designed to detect AI-generated or manipulated visuals even after cropping or compression.
The incident highlights growing challenges around trust and verification as artificial intelligence becomes more deeply embedded in the food delivery ecosystem.
In a separate case reported by TechCrunch, delivery platform DoorDash recently banned a rider for submitting an AI-generated image to fake a completed delivery.
