Rippling faces scrutiny over use of Chinese AI models

San Francisco-based HR startup Rippling is facing security concerns following reports that it utilizes models from Zhipu AI (also known as Z.ai). Zhipu AI is currently on a U.S. government entity list due to its connections to the Chinese military.

Security experts have raised alarms regarding the potential for data exfiltration and intelligence collection when sensitive information is processed through APIs connected to Chinese servers. Ryan Fedusiak of the American Enterprise Institute noted that the risks associated with AI models may exceed those of TikTok, as users provide highly sensitive personal and professional data to these systems.

Rippling, which services over 20,000 U.S. companies, has not responded to requests for comment regarding its use of Chinese AI or its data protection protocols. The trend of adopting Chinese AI models is being seen across various sectors; companies like DoorDash and Airbnb have also utilized Chinese models, such as Alibaba’s Qwen, citing high performance and lower costs compared to American alternatives like OpenAI and Anthropic.