Meta is using a new LLM trained to detect "seemingly benign ad content that is strongly suspected of covertly directing people to illegal content or harmful activity."
Read the full article on MediaPost →