Recently, a post on the social media platform Reddit accusing a food delivery platform of fraud has attracted widespread attention, but the twist came quickly. According to AIbase, this so-called "internal exposé" claiming that the company exploits drivers and steals tips was ultimately proven to be entirely fake information generated by AI.
The post initially used the tone of a whistleblower, vividly describing how it exposed the company's misconduct while drunk using public Wi-Fi. The content was highly inciting and quickly received over 87,000 likes, spreading widely to other platforms such as X. However, a senior technology journalist uncovered inconsistencies during the verification process. AIbase found that the so-called employee ID photo and an 18-page "technical document" provided by the poster showed clear signs of AI forgery.
More dramatically, the one who identified this scam was also AI. According to the report, investigators used Google's Gemini and SynthID watermarking technology to successfully confirm that the relevant evidence images contained digital watermarks from AI-generated content.
Key Points:
🛑 Fast Reversal: A "food delivery platform fraud expose" that gained tens of thousands of likes on Reddit was confirmed to be a fake information fabricated by AI.
🤖 Upgraded Forgery Techniques: The poster not only used AI to generate text but also forged misleading employee IDs and an 18-page fake technical document.
🔍 AI Identifies AI: This investigation used Google's Gemini and its SynthID watermark detection technology to finally expose this scheme of poisoning online public opinion with AI.
