Online bullying rising again; Kids using AI to humiliate each other, teachers

Dutch expertise center Offlimits has reported a surge in online bullying involving AI-generated or manipulated images, with schoolchildren using the technology to humiliate peers and teachers. Between January and September 2026, Offlimits received 955 reports of online bullying or abuse, already exceeding the 860 reports recorded for all of 2025 and more than triple the 298 reports from 2024. The organization noted a specific increase in cases involving AI-generated or manipulated visual content deployed as weapons of harassment in school environments.

Because this category of abuse involves synthetic or manipulated imagery rather than authentic photographs, it creates layers of harm that existing image-based bullying filters may not reliably detect. Guardii addresses exactly this threat vector through its anti-cyberbullying and AI-generated CSAM detection modules, both designed to surface escalating harassment patterns before harm crystallizes. Rather than requiring manual review of every image shared, the system identifies sustained targeting, escalation of hostility, and coercive dynamics across direct-message threads on Instagram, Snapchat, Discord and other platforms, flagging a child in distress to a parent or school authority in real time while preserving privacy for all parties.