Initiative M is an Artificial Intelligence project that uses machine learning to train a system to recognize instances of bullying on social media websites. An automated browsing agent browses posts on Twitter, Instagram, Tumblr, and other platforms to detects bullying and report the incidents.
Initiative M can monitor the profiles of your children unintrusively
Your children value their privacy online, even from you. So how do you monitor conversations and make sure they are not being abused and intimidated? Initiative M is a computer system that doesn't understand what is being said and never saves conversations. Yet, it can detect verbal abuse and bullying and can notify you in case intervention is needed. This way, you are not creeping profiles or intruding on privacy, and you still safeguard your children. Initiative M can access any public profiles or private profiles that you give it access to.
We ran a version of Initiative M in August, having it monitor over 1 million random profiles across Twitter and Instagram. During this period, it flagged 7,309 users as being bullies. To verify these results, we anonymized them, split up the findings into 10 equal sets and gave each set to 3 school teachers to verify that they were in fact instances of bullying. Of all the bullies flagged by Initiative M, only one of them was not confirmed as a bully by one of the teachers. While the results are excellent, we are not able at this point to validate that there were no other bullies that were not detected by the system.
Initiative M is still in a testing period. We hope to have a functional and accurate release available in June 2019. If you would like to be notified about the availability of Initiative M, please provide your email address below and we will contact you when it is publicly released. Your email will not be shared and will be used only for the purpose of notification.