Chinese AI creation platform LiblibAI issued an apology on Tuesday and vowed to strengthen content safety management, after state broadcaster CCTV News reported that its app generated a semi‑nude dancing video without triggering content controls. According to a statement released on its official WeChat account, the company said that it has taken the issue very seriously and immediately launched an internal review.

"We sincerely apologize," the company said in a statement. "The issue was promptly resolved with a technical fix, and all potential risk paths have been comprehensively blocked."

On April 12, CCTV News reported that a reporter using LiblibAI entered prompt phrases purchased online. Within minutes, the app generated nude woman dancing. Throughout the entire generation process, no platform restrictions were triggered, and the violating video was produced successfully.

According to the report, in addiction to AI platforms that generate sexually suggestive or pornographic videos via prompt-based questions-and-answer interactions, there are also dedicated software products on the market that profit specifically from using AI to create pornographic videos.

LiblibAI is an AI-native company that develops multimodal generative tools.

The company stated in the statement that it has taken note of recent media discussions on AI‑generated content safety and launched a comprehensive self‑inspection of its generation and review mechanisms.

The company said investigations found that in certain edge cases involving complex prompt combinations and evasive phrasing, the platform could produce content that did not meet regulatory standards.

"We deeply apologize," the statement said, adding that the incident exposed shortcomings in the platform's boundary recognition and governance mechanisms.

LiblibAI said it had implemented technical fixes immediately and closed the identified risk pathways. The company also pledged to strengthen red‑team testing, enhance review procedures, improve detection and response capabilities, and carry out an internal accountability review to refine management and moderation processes.

Stating that "content safety is the platform's bottom line," LiblibAI called on users and the public to assist monitoring the platform while it upgrades safety systems and works toward a healthier online ecosystem.

Global Times