Researchers say more than 250 ads containing AI-generated child sexual abuse material have appeared on Meta's platforms including Facebook and Instagram since early August, weeks after the company removed dozens of similar ads following reporting from Wired.