The Internet Watch Foundation warns that generative AI is being used to create child sexual abuse imagery, and finds ~3K AI-made abuse images breaking UK law
Internet Watch Foundation finds 3,000 AI-made abuse images breaking UK law — The “worst nightmares” about artificial intelligence …
Context & Ripple Effects
This finding turned an earlier expert warning about generative AI appearing in CSAM communities into a documented UK enforcement problem. It also extends the Internet Watch Foundation’s existing work on hashing known abuse images to prevent their reuse into a category where imagery can be newly generated rather than simply recirculated.
Later coverage shows the issue shifting from still images toward a wider and faster-moving format mix: the Foundation identified a sharp rise in AI-generated abuse videos in the first half of 2025 and later reported thousands of realistic AI-generated images and videos.
First-order effects
- The finding establishes that AI-made abuse imagery was already crossing the UK legal threshold, giving the Internet Watch Foundation a concrete new class of material to identify and report.
- Child-safety enforcement teams must contend with synthetic material that can be produced anew, reducing the usefulness of workflows focused only on previously known files and hashes.
Second-order effects
- The evidence strengthens the case for generation tools and hosting services to add prevention, reporting, and detection controls for abusive synthetic content; UK lawmakers later made AI tools designed to produce CSAM illegal.
- Detection systems will need to combine provenance and behavioral signals with hash matching, because novel AI-generated files may not match a pre-existing abuse-image database.
Third-order effects
- If synthetic CSAM continues to diversify into video and other realistic formats, child-safety enforcement will increasingly become an AI-governance and product-design issue, not solely a content-removal task.
- The likely durable shift is toward rules that cover the creation tools and distribution pathways as well as the illegal outputs, though the effectiveness of that approach depends on enforceable detection and reporting capabilities.
The trend: Generative AI is expanding the child-safety enforcement surface from identifying known illegal media to governing tools that can create new illegal content at scale.