Recent criminal prosecutions involving AI-generated indecent images of children have confirmed what safeguarding professionals have feared: deepfake abuse is no longer hypothetical.