DeepSeek

Our age is 16. Caution because of a chatbot with no teen settings or parent controls, safety limits that testers got past, and chats stored in China.

Parent guide checked November 2026.

DeepSeek at a glance
Our age
16+
Our rating
Use Caution
Chat
Text chats and file uploads with DeepSeek's AI, not with other people.
Spending
None. The app is free, and neither store lists ads or in-app purchases.
Parent controls
None built in. Use your phone's Screen Time or Family Link.
Plays on
iOS, Android, Web
Users
Unknown (100M+ Android downloads)

Why we say caution

DeepSeek is a free AI chatbot made by Hangzhou DeepSeek Artificial Intelligence Co., a company based in China. Kids type a question or upload a file and it answers: homework help, writing, coding, explanations or just conversation. It can search the web and has a "DeepThink" mode for harder problems. The app arrived on the App Store and Google Play in January 2025 and has been downloaded more than 100 million times on Android. There are no friend lists, feeds or messages between users. DeepSeek's terms say the service is "primarily intended for adults" and that anyone under 18 should use it only with a parent or guardian's consent. There is no teen mode and no parent account.

  • Data Stored in ChinaDeepSeek's privacy policy says it collects and stores users' personal data in China, and that includes what your child types and the files they upload. Chats can also be used to improve DeepSeek's AI unless "Improve the model for everyone" is switched off.
  • No Teen ProtectionsThere is no teen mode, no parent account and no way for a parent to see chats. The App Store listing says some content may be restricted through an age check in the app, but DeepSeek's own terms and privacy policy don't describe one.
  • Safety Limits That Give WayTesters got DeepSeek to write sexual role-play and to help with planning violence. In September 2025, the US government's National Institute of Standards and Technology (NIST) found a DeepSeek model answered 94% of plainly harmful requests when a common trick was used to get around its rules.
  • Slanted or Wrong AnswersNIST found DeepSeek's models repeated four times as many inaccurate and misleading Chinese Communist Party narratives as the US models it compared them with. DeepSeek's terms say its answers may contain errors and are for reference only.
  • Security LapsesIn January 2025, the security firm Wiz found a DeepSeek database open to anyone online, holding more than a million log entries that included chat history. DeepSeek closed it promptly once told.

Talk about it

  • Have you or your friends tried DeepSeek or other AI chatbots? What do you use them for?Starts from curiosity, not suspicion.
  • DeepSeek keeps what you type on servers in China. What would you never type into any chatbot?Builds the habit of keeping personal details out.
  • Has a chatbot ever told you something that turned out to be wrong, or only gave one side?Builds the habit of double-checking.
  • If something was really bothering you, who would you want to talk to?Reminds them that people, not a chatbot, are there for the hard stuff.

Settings to change first

DeepSeek has no parent settings, so use the phone's own controls. For a child under 16, block it by name: on iPhone, Settings > Family > your child > Apps & Websites > DeepSeek > Block App; on Android, Family Link > Screen time > Time limits > App limits. Block chat.deepseek.com in the browser filter too, because DeepSeek also works on the web. If your older teen uses it, switch off "Improve the model for everyone" together, keep names, school, photos and other personal details out of chats, and check important answers against another source.

The complete DeepSeek parent guide, with age-by-age settings, is part of the Family plan.