JP Morgan develops own algorithms to identify hateful content on YouTube after finding Google’s own filters inadequate

American banking giant, JP Morgan, has embarrassed Google by developing a system for stopping its ads appearing alongside hate video on Google’s YouTube platform that is better than Google’s own solution.

JP Morgan pulled its advertising from YouTube in March 2017 after it discovered that its ads were appearing alongside inappropriate content.

Google-owned video sharing website YouTube has come under criticism from politicians and advertisers after American banking giant JP Morgan develops its own software to remove advertising from questionable material on the platform.

JP Morgan’s Executive Director of Media and Channel Strategy, Jake Davidow, said: “When news broke about ads finding their way next to horrific pieces of content, we paused our efforts and pulled our ads from YouTube.”

The bank has since developed its own algorithms that use YouTube’s own application programming interface (API) to filter channels on the platform based on comments, subscribers, video count, topics, language and more, using seventeen different filters in total.

Since using the algorithms, JP Morgan has identified 3,000 “safe” channels out of a potential five million on YouTube.

Since developing the technology, JP Morgan has reinstated its advertising with YouTube but the matter has generated controversy with journalists questioning why the company had to take matters into its own hands, highlighting the shortcomings of Google’s own safeguarding systems.

Campaign Against Antisemitism is currently collaborating with Google in efforts to clean up antisemitic hatred on YouTube. We are currently looking for a team to help us flag antisemitic videos using special tools provided by Google. If you would like to join the team, please go to antisemitism.org/volunteer.

Image credit: Gideon Benari