AI on Friendstime: translation, suggestions and FriendsBot
Friendstime uses artificial intelligence in a few places: for translating, for suggestions while you write, as the chat assistant FriendsBot and for a fact check of public posts. To do so, Friendstime sends text or images to the AI service Google Gemini. This page shows where that happens and how you control it in the Android app.
Where Friendstime uses AI
Friendstime uses Google Gemini as its AI service. These are the places: your chat with FriendsBot, translating posts, comments and single chat messages, hashtag suggestions for your draft, caption suggestions for your picture, the fact check of public posts and, if switched on, the check of post images. Reply suggestions exist only in the FriendsBot chat. In the Android app you turn the AI features off under Settings with “Allow AI features”; the translate buttons under posts, the suggestions in the post window and FriendsBot then disappear.
Translating posts and comments
Under posts there is a “Translate” button, on the web from 12 characters of text, in the Android app from 15; on the web you can also translate comments. The target language is the language of your browser or device, and a second tap shows the original again. Text is translated only when you tap, and at most 5,000 characters at a time.
Translating chat messages
Single chat messages can be translated too, but only unencrypted ones: for end-to-end encrypted messages the menu item is missing. When you tap “Translate”, the text of that one message goes to Friendstime and from there to Google Gemini; the app tells you so first. Friendstime does not store translations of chat messages.
Suggestions while you write
In the Android app the AI suggests 3 to 5 hashtags for your text in the post window when you ask for them; the text needs at least 5 characters. For a picture there are three caption suggestions: a factual, a funny and an emotional one. Both start only when you tap the matching button.
FriendsBot
FriendsBot is an AI assistant in the chat list of the Android app. It answers in German, briefly and in a friendly tone. Your messages to it go to the AI and are not end-to-end encrypted; the app shows reply suggestions only in this chat.
Fact check of public posts
If you publish a public post of at least 50 characters on the web or in the Android app, the AI checks its claims after you send it. The result appears as a note on the post, for example “Fact-checked”, “False information”, “Context needed” or “Disputed”. Posts for friends only, close friends only or just you do not go to the AI. An AI check of post images runs only if it is switched on in the admin area.
Try it yourself
Signing up takes a minute and is free.
Frequently asked questions
Which AI service is used?
Friendstime uses Google Gemini. The text or image in question is sent to the service for this; how Google processes inputs is governed by the provider’s own terms.
Do my chats go to the AI?
Not on their own. Apart from your messages to FriendsBot, only single unencrypted chat messages go to the AI, namely when you tap them yourself to translate. End-to-end encrypted messages cannot be translated.
Does the AI check my private posts?
No. The automatic fact check applies to public posts only. Posts for friends, close friends or only you do not go to the AI.
What is on the web and what is in the Android app?
Translating posts is in both, comments are translated on the web, and the automatic fact check runs for public posts from the web and from the Android app. Hashtag and caption suggestions, FriendsBot and the “Allow AI features” switch are in the Android app.
Is every image checked by the AI?
Only if the AI image check is switched on in the admin area. Caption suggestions, by contrast, you get only when you request them yourself.
Do I have to agree to the AI?
In the Android app a notice with “Allow”, “Decline” and the policy appears the first time you tap an AI feature. If you decline, the AI features are off; later you can allow them again in the settings.