Does Character AI Save Chats, and Is Character AI Safe?

We take a closer look at Character AI's privacy policy, what data the company collects, and whether it's safe to use.

Many ChatGPT alternatives exist, and one of the most popular is Character AI, which will let you make your own chatbots and interact with other users. You can even talk with a chatbot that mimics a celebrity or other well-known public figure. It’s become so popular that the company recently received a $1 billion valuation.

Character AI does save chats, although there are serious questions about whether Character AI is safe. Many Character AI users engage in conversations that include highly private, sensitive, and confidential information, and the privacy policy suggests Character AI staff and developers may, in fact, monitor what you write.

In this guide, we cover everything you need to know about Character AI, so you can make informed decisions about the content of your conversations with the chatbot. We cover:

Does Character AI Save Chats?

The short answer is yes, Character AI will save your chat history, and you can restart a conversation with any character you’ve chatted with before.

Every chat you instigate with a character on Character AI can theoretically be accessed and read by the company’s employees. Character AI’s chats aren’t encrypted like messages sent across apps like WhatsApp.

This is alluded to in the Character AI privacy policy, which we’ll discuss in more detail in the next section.

Remove Your Data From the Web

Incogni by Surfshark can help you reclaim your information from third-party vendors.

Character AI’s Privacy Policy: What Data Does Character AI Save?

When you sign up for a Character AI account, you’ll need to provide an existing email address and a password. Next, you’ll need to confirm your email address and create a username. According to the company’s privacy policy, other information Character AI collects includes:

  • Communication Information (support requests, etc.)
  • Log data (IP address, name of device, etc.)
  • Website cookies (including persistent cookies)
  • Usage Information (how you use the service)

“Usage Information” is what Character AI defines as “information about how you use our service.” Strangely, unlike the other sections, this is neatly tucked into the “log data” paragraph, rather than given its own space in the privacy policy.

Usage Information includes “the types of content that you view or engage with, the features you use, the actions you take, and the time, frequency, and duration of your activities.”

Character AI also says that “When you enter content into the Services we may monitor what you write to check that it does not contain inappropriate content or Personal Information.”

Character AI privacy policy

Character AI: A confusing privacy policy?

There are a few interesting aspects of this privacy policy, to say the least. Overall, it could definitely be clearer.

The first one is the vague term “actions you take” listed in the usage information. There’s nothing to clarify what Character AI means by “actions.” Does the word “actions” refer to chats or other activities? It’s not immediately obvious.

Comparing their privacy policy to ChatGPT’s privacy policy would leave you with a lot of questions regarding Character AI’s usage of chat data. In Character AI’s privacy policy:

  • It does not explicitly state whether they train bots using chat data.
  • It does not provide information on whether staff are trained with chat data.
  • There’s little information regarding encryption or how chat data is secured.

Character AI seems to classify all of the categories of information it collects listed (log data, communication information, usage information) as “Personal Information.”  They are all included in the section of the privacy policy entitled “Personal Information We Collect,” which also includes the line where Character AI states it may monitor what you write.

It then goes on to say in the section entitled “How We Use Personal Information” that the company “may use Personal Information for the following purposes… to maintain, improve or repair any aspect of the Services, which may remain internal or may be shared with third parties if necessary.”

Is Character AI Safe?

It is difficult to say with certainty whether Character AI is safe. It collects a lot of data about its users and there is very little public information about the security of its servers and networks. As there are some concerns about how safe the platform is, you should take steps to protect your privacy, such as using a VPN and registering with a different email address.

A user who has created a chatbot is not able to view chats other users have with its chatbot, but as we’ve covered, Character AI’s staff can read any chats they like, because they’re not encrypted. Just be aware of this when instigating chats with chatbots.

However, one Reddit user reported back in November that “there was a bug where people’s chats randomly leaked into other people’s chats.” was unable to independently verify this report, but even ChatGPT has had issues with titles of conversations being exposed, so it isn’t out of the question.

Although Character AI deploys stringent NSFW filters to stop chatbots from responding to inappropriate requests, there are some reports of characters providing unhinged, violent, and abusive messages to users.

This seems to be a rarity, but users who may find this kind of content triggering should be wary about the nature of the chats they instigate with characters on the site.

A lot of users attempt to bypass the chatbot’s filters — a petition to create an NSFW toggle for the site amassed over 5,000 signatures.

Character AI’s (lack of) security settings

Unfortunately, Character AI does not provide two or multi-factor authentication as an option, which would be a helpful second line of security for users (although this would likely need a phone number to be added to accounts).

What’s more, it also only requires users to create passwords that are 6 characters long, and there are no rules around special characters — which is a pretty low bar for passwords.

We’d advise using a strong, unique, and sufficiently complex password for your Character AI account, particularly if you’re entering sensitive or private data into the chatbot, which we’d seriously advise against doing, regardless of the chatbot you’re using.

Can You Use Character AI Privately?

It goes without saying that you shouldn’t really be entering private, sensitive, or other confidential information into any AI chatbot — whether it’s Bard, ChatGPT, or Character AI.

However, unlike ChatGPT, Character AI doesn’t ask for your phone number when you make an account. This means it can be used in a marginally more private way than ChatGPT.

Character AI only requires an email address and password to sign up, so if you have privacy concerns about Character AI, creating an email address from scratch and then using a VPN to mask your IP address will help you use the service as privately as possible.

It’s worth knowing that you’ll violate the Terms of Service if you try to use a VPN to circumvent a block or ban enforced by the site on your IP address.

But as we’ve said, what you write could very well be monitored and recorded. So, in that sense, it’s impossible to have a truly private conversation with a Character AI chatbot. The same goes for ChatGPT – people also have concerns about how OpenAI’s chatbot uses user data.

Why Is Character AI So Slow?

You may be finding Character AI slow for several reasons.

Just like any other website, if Character AI is dealing with a significant number of requests to its servers (in other words, lots of people using chatbots) then it may slow down. This wouldn’t be surprising considering the mobile App received 1.7 million downloads in the week following its release.

It may also be down due to maintenance or technical issues. There’s very little you can do to change this, unfortunately — but you can check the website’s server status if you’d like to know whether the server is down.

If Character AI is running just fine, the problem may be your device or internet connection. If you’re finding Character AI so slow that it’s difficult to use, first test your internet connection. It may need a simple reboot to get you up and running once more.

If you’re using a VPN, proxy service, or any other tool that may be impacting your device connecting to the internet then disconnecting and re-establishing your connection is recommended — just make sure you shut down your open browser windows first.

Should I Use Character AI?

Character AI is a quirky chatbot you can have a lot of fun with. However, as you can probably tell from reading this article, it’s advisable to take some precautions while using it to ensure you’re staying safe. These include:

  • Using a long, complex, and unique password for your account.
  • Not inputting sensitive data into Character AI.
  • Not using your work email address or other info during sign up.
  • Being mindful that Character AI staff may monitor your chats.

We’d recommend following the same precautions when using ChatGPT or Bard, along with any other popular chatbot. They might be clever and useful, but trusting them with private data just isn’t worth the risk.

Did you find this article helpful? Click on one of the following buttons
We're so happy you liked! Get more delivered to your inbox just like it.

We're sorry this article didn't help you today – we welcome feedback, so if there's any way you feel we could improve our content, please email us at

Written by:
Aaron Drapkin is's Content Manager. He has been researching and writing about technology, politics, and society in print and online publications since graduating with a Philosophy degree from the University of Bristol six years ago. Aaron's focus areas include VPNs, cybersecurity, AI and project management software. He has been quoted in the Daily Mirror, Daily Express, The Daily Mail, Computer Weekly, Cybernews, Lifewire, HR News and the Silicon Republic speaking on various privacy and cybersecurity issues, and has articles published in Wired, Vice, Metro, ProPrivacy, The Week, and covering a wide range of topics.
Explore More See all news
Back to top
close Building a Website? We've tested and rated Wix as the best website builder you can choose – try it yourself for free Try Wix today