TikToker Admits to Using ChatGPT to Edit DoorDash Order Photos for Refund
A TikToker used ChatGPT to edit food photos, making chicken appear undercooked to strengthen her refund claim, ultimately receiving a $39.24 refund.
She publicly shared the process on TikTok and thanked ChatGPT, with the video garnering over 4.2 million views. DoorDash responded in the comments, saying "we should be blocked."
This incident has sparked widespread controversy, being labeled a typical case of AI-assisted food refund fraud.
Source: Public Information
ABAB AI Insight
The TikToker (@mi5under5t00d) incident occurred recently with a DoorDash order where she first received cold chicken and missing sides, then used ChatGPT to generate more obvious "raw chicken" images submitted to customer service. Similar AI-edited food photo refund fraud cases began to emerge and spread by the end of 2025.
In terms of capital pathways, delivery platforms like DoorDash are shifting resources from manual customer service reviews to investments in AI image detection and anti-fraud systems, motivated by the need to control the growing losses from AI-assisted refunds while protecting the interests of drivers and restaurants, and preventing genuine complaints from being overlooked due to the proliferation of false evidence.
The trend of AI photo editing for refunds has also appeared concurrently with platforms like Uber Eats and Grubhub, as the delivery industry is transitioning from traditional manual judgments to AI image authenticity verification. DoorDash's public response marks the platform's proactive counteraction.
This fundamentally represents a shift in technology and pricing power: consumer refund evidence is shifting from real photos to AI-generated/edited images, with tools like ChatGPT significantly lowering the barrier for forgery, making it difficult for platform customer service to discern authenticity. In the long term, this will force delivery platforms to shift pricing power from lenient refund policies to strict AI anti-fraud verification, while accelerating capital concentration from manual customer service to image authenticity verification infrastructure.
ABAB News · Cognitive Law
When AI lowers the cost of wrongdoing, trust becomes the scarcest resource. Convenience may feel good temporarily, but when platforms retaliate, it leads to social death online; openly flaunting scams is akin to walking into a trap. When AI can turn cooked food into raw meat, the platform's pricing power shifts from user-friendly to anti-fraud defense.