Text-to-image models have made it significantly easier to generate synthetic images, including harmful and illegal content such as child sexual abuse material (CSAM). In public debates and policy discussions, a commonly proposed safeguard is to remove images of children from training datasets.
Dataset filtering provides only limited protection against CSAM generation in text-to-image models, study finds
Reader’s Picks
-
South Africa is pioneering new ways to embed ethical benefit sharing in genomics research through community-led decision-making. Speaking at the [...]
-
Mass shooting incidents and car crashes may seem like two unrelated incidents, but a recent study has uncovered that there [...]
-
Framing environmental risks in terms of how much time is left, rather than a future date, makes them feel more [...]