Cyberbullying has been an ongoing problem for the social media platform Instagram and Adam Mosseri, head of the company, says that he would much rather have lesser users than let cyber bullies continue spreading negativity.
To try and solve the issue, Mosseri announced that Instagram will be upgraded with new tools to stop bullying from spreading instead of blocking posts or bullies from the platform. One of these will incorporate artificial intelligence (AI) to recognize offensive comments and another will be able to inhibit select accounts from publicly commenting on their posts.
Mosseri says that he will roll out these tools in the hopes that they will “prevent bullying from happening on Instagram” and “empower the targets of bullying to stand up for themselves”.
Think before you act
The Rethink tool will use AI to spot offensive comments and give users a chance to rethink their comment by posing a question in front of them: “Are you sure you want to post this?”
Instagram suggests that research has shown that this is successful in making “some” users “undo their comment and share something less hurtful.”
“Some” isn’t very reliable, which is why there is another tool in store for comments that aren’t picked up by the AI filter.
The power to shadow ban
According to Mosseri, targets are often afraid to block a user who has passed an offensive comment because of the consequences involved, which is why a second tool called Restrict will be rolled out. This will allow targets to block such users without them finding out.
These restricted users will still be able to see future posts from the targets, but will not know whether the target is online or has seen their posts. Targets will also be able to review comments from the bullies before they are posted, giving them the choice to make the comments public, delete them, or leave them pending.
Such comments will be concealed behind a “sensitivity screen” that will have to be pressed to be viewed, hence if targets don’t want to see the comments, they won’t need to.
Direct messages from restricted users will not be viewable by targets, nor will they be notified of them. These messages will be sent to spam where they can be viewed only if the targets un-restrict the user first.
This article originally appeared on Tech Radar.