Bullying affects everyone: those who are bullied, those who bully others, and those who witness bullying. And it doesn’t stop in the physical world. It happens online too.
Kindly is an open-source anti-cyberbullying API (application program interface) powered by artificial intelligence and natural language processing that gives children themselves the power to eliminate cyberbullying and center their own safety and protections in the solution.
Kindly aims to end cyberbullying, one message at a time. It evaluates messages before they are sent and offers users the chance to modify them.
Test Kindly for yourself. Reply to the test message on the site and check your entry with Kindly to see if your message is toxic or considered cyberbullying. Feedback on the message offers users the chance to edit their original message. After seeing and/or responding to the feedback, they can choose whether to send the final message.
Users can also help build and improve the machine learning model powering Kindly by contributing to the training dataset. After testing a string of text with the tool, indicate if your entry contained cyberbullying or toxic intent. This will help improve Kindly’s understanding, and you’ll be part of the solution to cyberbullying.
Kindly was first conceptualized by a 16-year-old young innovator/youth advocate, Gitanjali Rao, TIME’s first-ever Kid of the Year. The tool is the product of a collaboration between Rao and UNICEF. It is now recognized as a UNICEF “Digital Public Good” and accessible worldwide.