Comment by myrryr

Wow, the false positive rate on that will be interesting. from OpenAI's point of view, having it flood the parents with false positives is _way_ safer than having some slip past.

so needless to say....