The content in question was actually something they deemed could potentially be depicting something that would be non-consensual. When we reviewed it, the content should not have made it on the site. It actually made it up through an edge case by making it through not only three of our softwares, but two other softwares that we actually had that were running against the databases. Since then, we've plugged it.
We understand, and it goes back to what we said: Even one example is too many, and that's why we constantly work to identify these edge cases and plug them, so they can never happen again.