Data Protection Day is today, Jan. 28, 2025, and it’s a world event intended to help people remember to follow the best practices for data privacy and security. It’s only a coincidence that DeepSeek, a new and free AI chatbot with an extremely questionably privacy policy, vaulted to the top charts of the App Store and Google Play Store shortly before Data Protection Day began.
Chances are, you’ve heard the name DeepSeek over the last day or two. Within the tech industry and stock market, the attention around DeepSeek AI has been more about the open-source nature of its models and the cost required to train them rather than their effectiveness compared to chatbots like ChatGPT or Gemini. “Despite its strong performance, it also maintains economical training costs,” DeepSeek claims in a white paper.
The short recap is that DeepSeek’s latest open-source AI model, DeepSeek V3, quickly became lauded by researchers and benchmarks. It performs comparably to the top OpenAI models publicly available, despite the fact that DeepSeek V3 is free for users to access and is supposedly cheap to train.
“Comprehensive evaluations demonstrate that DeepSeek-V3 has emerged as the strongest open-source model currently available, and achieves performance comparable to leading closed-source models like GPT-4o and Claude-3.5-Sonnet,” wrote DeepSeek researchers in the conclusion of the white paper. The company’s internal findings have since been supported by independent tests, hence why the DeepSeek AI app has blown up globally.
So, why are users mass downloading DeepSeek, and why should that give you pause on Data Protection Day?
For starters, it’s worth noting that much of the fear in the U.S. about Chinese-based apps and services is relatively unfounded. As an example, TikTok stores all U.S. user data in a secure Oracle cloud environment — it’s not stored in China, and it’s not even stored in TikTok’s international servers in Virginia or Singapore.
It’s the opposite with DeepSeek. To its credit, the company is being transparent about its data storage practices and privacy policy. It’s all outlined on the DeepSeek website, but it’s not something you should be eager to agree to.
“The personal information we collect from you may be stored on a server located outside of the country where you live,” the company explains. “We store the information we collect in secure servers located in the People’s Republic of China.”
This is problematic due to China’s Cybersecurity Law and National Intelligence Law, which require companies in China to share data with the government when requested. Being that DeepSeek AI is based in China and all user data is stored there, it is completely possible that the company could share your information with the Chinese government and other companies.
In fact, a Wired report found that DeepSeek AI already appears to send data to Baidu, a Chinese technology giant specializing in search, and Volces, a Chinese infrastructure company.
Contrary to popular belief, DeepSeek’s ties to China are not the reason you shouldn’t use the app. It’s a factor, sure, but the privacy concerns are much simpler than that. You shouldn’t be willing to give up the data DeepSeek admits it collects to any private corporation or government, whether it be in the U.S., China, or any other region.
Simply put, DeepSeek collects a lot of data. It might know your birthday, email address, or phone number, depending on how you sign up. If you use an Apple or Google account, it collects some information from those accounts, like access tokens.
More worryingly, the company receives data that has nothing to do with DeepSeek AI. Using the information you gave DeepSeek, it tries to match its profile about you to your actions elsewhere. The official DeepSeek privacy policy outlines everything:
“Advertisers, measurement, and other partners share information with us about you and the actions you have taken outside of the Service, such as your activities on other websites and apps or in stores, including the products or services you purchased, online or in person. These partners also share information with us, such as mobile identifiers for advertising, hashed email addresses and phone numbers, and cookie identifiers, which we use to help match you and your actions outside of the Service.”
If that wasn’t enough to spook you, DeepSeek also collects your keystroke patterns or rhythms, as well as your IP address. In other words, the company registers what you’re typing, even if it’s in another app or if you never enter your DeepSeek prompt.
To be fair, consumers don’t have a ton of leverage when it comes to using products and services in the U.S. That’s why many of us simply accept the terms and conditions and move on. However, on Data Protection Day, DeepSeek gives us a great example of why we shouldn’t do that. You’re surrendering a lot more data than you probably thought you would by downloading a top-charting AI chatbot from the App Store or Google Play Store.
More importantly, apply scrutiny of companies’ privacy policies and data collection practices to more than just DeepSeek. No matter what U.S. politicians would lead you to believe, there’s plenty more to worry about than just Chinese companies. Google, OpenAI, Microsoft, Apple, and many more tech companies are all collecting varying levels of data from you, and it’s worth thinking about what you’re willing to share to access the latest AI models.