The new slate of features and settings rolled out in the Safety Mode are designed to protect users from abusive replies or unwanted interactions on Twitter.
Safety Mode has been rolled out in testing in the beta phase amongst a small group of users on Twitter who have English-language settings enabled, on iOS, Android, and Twitter.com. The mode will go through a gradual rollout as they gather feedback through the beta phase.
Twitter collaborated with partners with expertise in online safety, mental health, and human rights, including members of the Trust and Safety Council, and conducted listening and feedback sessions with them to develop the new safety mode.
The partners also played a role in selecting the feedback group, prioritizing people from marginalized communities and female journalists, to be a part of the beta phase. When enabled, Safety Mode will automatically block an account for seven days when these accounts repeatedly send uninvited replies or unwanted mentions, and when they use abusive or hateful language.
Accounts detected by Twitter systems to be potentially harmful or unsolicited will be unable to follow the user being (potentially) harassed, and these accounts will not be able to see the user’s Tweets, follow them, or send them Direct Messages.
The systems that detect potentially harmful or insulting behavior will also evaluate the Tweet’s content and the relationship between the Tweet author and replier. So accounts followed or frequently interacted with will not be auto-blocked.
When Authors of Tweets found by our technology to be harmful or uninvited will be auto-blocked, meaning they’ll temporarily be unable to follow your account, see your Tweets, or send you Direct Messages.
All information related to accounts flagged through Safety Mode can be found in the tab, and users can view the details. Safety Mode autoblocks can be seen and also undone through Settings. When the temporary block period ends, users will receive a notification summarizing the details of the block.
While general users are also prone to abusive replies, the new mode will be particularly beneficial for public personalities who are vulnerable to targeted harassment and trolls, due to the public nature of their profession. The mode also adds to the collective experience of being protected on the platform with features previously launched, such as conversation settings.