A Chinese AI app called Haotian has generated millions by selling ultra-realistic face-swapping technology on Telegram, primarily to scammers. The tool enables nearly perfect swaps during live video chats and integrates with apps like WhatsApp and WeChat. Its main channel disappeared after inquiries revealed its use in Southeast Asian fraud operations.
The artificial intelligence application Haotian, developed in China, specializes in advanced face-swapping capabilities that have proven highly effective for deceptive purposes. According to analysis, the app allows users to create "nearly perfect" face swaps in real-time video conversations, earning its creators millions of dollars through sales on the Telegram platform.
Haotian's technology stands out for its versatility and ease of integration with popular messaging services such as WhatsApp and WeChat. Users can fine-tune up to 50 different settings, including adjustments to cheekbone size and eye position, to closely replicate the appearance of the person they aim to impersonate. This level of customization makes it a powerful tool for various applications, but investigations have highlighted its darker side.
Researchers and independent reviews, including those conducted by WIRED, indicate that Haotian has been actively marketed to individuals involved in "pig butchering" scams—a type of romance fraud prevalent in online interactions. These operations, largely based in Southeast Asia, exploit the app to build false trust with victims before extracting money, often through cryptocurrency schemes. The app's promotion within fraud networks underscores the growing risks posed by accessible deepfake technologies.
Notably, the primary Telegram channel for Haotian vanished shortly after inquiries into its scammer connections, suggesting an attempt to evade scrutiny. While the platform offers robust features for legitimate creative uses, its documented ties to criminal activities raise concerns about the ethical deployment of AI in everyday digital communications. This incident highlights the challenges in regulating tools that blur the line between reality and deception in online spaces.
