The encyclopedia · Advertising & PR · Marketing decision · 2025
Lambency Detailing got ChatGPT to write fake 5-star reviews for its own cars
Car-detailing brand Lambency posted ChatGPT-written fake reviews using real customers' details. Singapore's CCCS called it its first AI-fake-review case.
Quantum Globe · 2025-07-03
What happened
Quantum Globe, the Singapore company behind the Lambency Detailing car-care brand, posted fake five-star reviews on its Sgcarmart business page. It used real customers' names, vehicle registration numbers and photographs — without their knowledge or consent — and generated the review text with ChatGPT.
The reviews were submitted through a QR code Sgcarmart provides (no account needed) and customised by ChatGPT to reflect the service each customer was supposed to have received. The practice ran for about two years.
Singapore's Competition and Consumer Commission ruled it the country's first case of AI-generated fake reviews, with its CEO describing the conduct as a 'dark pattern'. CCCS's algorithms had flagged suspicious clusters of five-star reviews posted on the same dates.
Quantum Globe and its director, Matthew Lim Jun Ho, each gave an undertaking to stop. The company agreed to remove the fake reviews within eight working days, set up a six-month customer feedback channel, notify affected customers, and publish corrective notices for six months.
Why it happened
- A QR code built for genuine happy customers became a machine for manufacturing their approval
- Using real names and registration numbers made fabricated praise look authentic
- Being the first firm caught using AI to fake reviews is a lead no business wants
The lesson
A tool that makes it easy to leave a review also makes it easy to fake one. Generating fabricated praise with AI and attaching real customers' details turns a platform's trust signal into a lie.
Aftermath
Singapore's CCCS ruled Quantum Globe had posted fabricated five-star reviews on Sgcarmart — using customers' names, registration numbers and photographs without their consent, text generated by ChatGPT over about two years. Its algorithms had flagged suspicious clusters of same-date five-star posts. Both the company and its director Matthew Lim Jun Ho gave undertakings: stop the fake reviews, remove them within eight working days, run a six-month feedback channel, notify affected customers, and publish notices for six months. CCCS called it the country's first AI-fake-review case.
Sources
spotted an error? The club wants to know.
More like this
OKQ8's 'Sweden's largest charging network with nearly 1,000 points' ruled misleading
BYD's skijoring SUV ad — skiing behind a skidding car — ruled reckless
Nissan's '80 kW fast-charging' Townstar ad upheld as misleading by the Dutch RCC
Somewhere, someone solved the problem this company failed at. 2nd Opinion →
Comments · 0
Sign in to join the comments.