MIT researchers developed a breakthrough way to detect AI models adapted to create child sexual abuse material without generating illegal content.
The technique could help platforms remove harmful models before they spread.
Want to know more?