"We know we have more work to do to further limit bullying and spread kindness on Instagram", he said. Instagram stressed that the new tools will help the company in ensuring the safety of teen users whom it believes experience online bullying most of the time.
According to Instagram, using this machine learning capability will enable it to "significantly" increase its detection of bullying content.
Instagram announced on its official blog Tuesday that it will be using machine learning to combat bullying on the platform. When encountering such instances, the AI technology will detect the photos and won't act as an executioner but instead, it will forward such uploads to the human reviewers who will review and flag the photos and videos as appropriate or inappropriate. Instagram also unveiled a "kindness camera effect" with dancer Maddie Ziegler of Sia's Chandelier (music video) fame where the users can post their photos with kind comments and hearts.
The tool - according to the company - will be more beneficial in removing bullying.
Instagram is toughening its approach against hateful posts with new artificial intelligence capable of detecting bullying in photos and their captions.More news: Minnesota Vikings regain control of season with key win over Eagles
More news: Rescuers still searching for missing child and two foreigners — MALLORCA FLOODING DISASTER
More news: Pipeline explodes near Prince George, B.C.: RCMP
The feature has already started rolling out and will continue rolling out over the next few weeks.
Users will also be empowered to select the comments that appear during their live videos, meaning they can hide and filter out any bullying messages. Your tagged friend will be notified as per usual, and they can then share this post to their own story, or "spread kindness" with a selfie of their own.
In May, the company started filtering out bullying comments meant to harass or upset people. When using the device's rear camera, the mode adds kind comments onto the image in various languages. This also includes comparison photos wherein users could post split-screen photos comparing someone in a negative way.