JP Morgan Chase has developed its own proprietary algorithm to ensure that its ads do not appear alongside unsavoury content on YouTube.

The bank pulled its ads from the platform last March after the brand safety scandal, only restarting its YouTube ads after developing and testing the tool. The algorithm, built by the company’s internal programmatic and media-buying teams, plugs into YouTubes API to select safe channels for its ads to appear on at scale.

In an interview, Jake Davidow, executive director of media and channel strategy at JPMorgan Chase, told Business Insider: “When news broke about ads finding their way next to horrific pieces of content, we paused our efforts and pulled our ads from YouTube. We wanted to figure out a scalable solution and make sure we got it right.”

Aaron Smolick, executive director of paid-media analytics and optimization at JPMorgan Chase, added: “The model that Google has built to monetize YouTube may work for it, but it doesnt work for us. The attention of protecting a brand has to fall on the actual people within the brand itself.”

The algorithm consists of 17 filters, which look for things like the total video count on a channel, (to sift out one-off viral videos); subscriber numbers; language; and comments left about other videos on the channel. It has enabled the bank to whittle down the number of YouTube channels on which its ads can run from 400,000 to just 10,000.