DeepSeek retains the right to gather information from a variety of external sources. For example, if you create a DeepSeek account using Google or Apple sign-in, the platform will obtain certain details from those companies. Additionally, advertisers provide data to DeepSeek, which could include “mobile identifiers for advertising, hashed email addresses, phone numbers, and cookie identifiers, all of which help us connect you and your behavior beyond the service,” as stated in its policies.
Utilization of Information by DeepSeek
A significant amount of data may flow to China from DeepSeek’s global user base, yet the company retains control over how it utilizes this information. According to DeepSeek’s privacy policy, it will use data in various standard ways, such as maintaining service functionality, enforcing terms and conditions, and enhancing the platform.
Importantly, the company’s privacy policy indicates that it may leverage user interactions to assist in developing new models. The company will “review, enhance, and develop the service, including by overseeing interactions and usage across your devices, analyzing user engagement, and training and refining our technology,” as per their policies.
DeepSeek’s privacy policy also mentions that information will be used to “fulfill our legal obligations”—a common clause found in many corporate policies. The policy notes that data can be accessed by its “corporate group” and shared with law enforcement and public authorities when necessary.
While all organizations have legal responsibilities, those operating in China face specific obligations. Over the last decade, Chinese authorities have enacted a series of cybersecurity and privacy laws allowing state officials to request data from tech firms. A law from 2017, for example, asserts that organizations and citizens must “cooperate with national intelligence efforts.”
These regulations, combined with escalating trade tensions between the United States and China and other geopolitical issues, have heightened security concerns regarding TikTok. Critics have argued that the app could collect vast amounts of data and send it back to China, potentially being used to disseminate Chinese propaganda (TikTok has denied transferring US user data to the Chinese government). Simultaneously, several DeepSeek users have highlighted that the platform fails to provide satisfactory responses to inquiries regarding the 1989 Tiananmen Square massacre and sometimes offers answers that appear propaganda-like.
Willemsen points out that in contrast to users on platforms like TikTok, individuals engaging with a generative AI system tend to be more involved, making the content feel more personalized. Therefore, any potential influence could be greater. “The risks of subtle content modification, steering conversations, and heightened engagement should logically provoke increased concern, not less,” he states. “Particularly given the opacity surrounding the model’s mechanisms, thresholds, boundaries, censorship protocols, and objectives/personas, all of which remain largely unexamined, and its considerable popularity at such an early stage.”
Olejnik from King’s College London mentions that while the TikTok ban was a particular instance, lawmakers in the US or other nations could potentially take similar actions in the future. “We cannot discount the possibility that 2025 could see an expansion: direct measures against AI companies,” Olejnik notes. “Naturally, data collection could again be cited as the reason.”
Updated 5:27 pm EST, January 27, 2025: Added additional details about the DeepSeek website’s activity.
