Tim Green
Leading AI video systems are generating hateful and extremist content in up to 40% of problematic prompts, exposing critical safety and ethical failures. Urgent action is needed to prioritise responsible development, regulation, and community-led safeguards.
Discover more at https://dev.to/rawveg/the-video-ai-hate-problem-5h9h
#HumanInTheLoop #AIGeneration #ContentModeration #AIethics
The Video AI Hate Problem

In October 2025, researchers at the Anti-Defamation League's Centre on Technology and Society...

DEV Community
Feb 14 at 12:00pmWeb
Show threadAlexander LothFeb 14
@rawveg This aligns with what we're seeing on the text side too. In our JudgeGPT research, participants struggled to distinguish LLM-generated news from human-written articles. The safety gap spans modalities. The 40% figure for video is alarming.