BCI warns of surge in AI‑generated child exploitation; lawmakers debate penalties, training and tech funding

State investigators briefed the Protection and Victim Services Committee on how generative AI is changing child exploitation investigations, presenting local case examples, statewide tip statistics and recommendations for funding and legal changes.

Special Agent Cassie Halsa, commander for internet safety against children at the North Dakota Bureau of Criminal Investigation, told lawmakers that AI has “made it faster, easier, and more scalable while making it harder for us to detect and to investigate.” She said North Dakota received 2,698 cyber tips in 2025 and that an increasing share involve AI‑assisted manipulation — using photos of real people, including minors, to create explicit images with few obvious signs of manipulation.

Cases and impact: Halsa described instances in which an offender used publicly available images and AI tools to produce explicit material, and another case where a juvenile’s likeness was used without her knowledge. She described school incidents in which AI‑generated explicit images of classmates were widely shared, causing therapy needs and, in some cases, students leaving school. “This child had no idea these images existed,” she said. Investigators urged treating possession and distribution of AI‑generated CSAM with the same seriousness as other CSAM because identifying synthetic content early is often impossible.

Investigation and resource needs: BCI said digital forensics must evolve — investigators need specialized software, cryptocurrency and dark‑web tracking tools, and continual training to keep pace with rapidly changing models. Halsa also urged sustained funding for immediate victim services; she said the ICAC task force model of bringing victim services on scene has been effective but underfunded.

Legal and policy debates: Committee members pressed on sentencing outcomes, plea practices (including Alford pleas), geofencing or tech‑blocking approaches, and whether state law should require disclosures or restrictions for commercial AI operators. Counsel described recent state code updates that added computer‑generated images into definitions and noted federal Supreme Court developments that affect enforcement and age‑verification rules. Members discussed legislative options including: (1) stronger minimum penalties or sentencing guidance, (2) civil remedies and state‑backed civil actions on behalf of victims, (3) mandatory annual education for judges and prosecutors, and (4) funding to equip and staff digital‑forensics teams.

What’s next: Lawmakers asked counsel to draft options for the committee to consider — from funding requests to targeted statutory changes — and flagged judicial and prosecutor training as a likely noncontroversial first step. No bill votes were taken at the meeting.