The Winning Solution:
The team developed Hate Shield — a responsive tool to detect and report online hate speech and harassment. The platform enables users to upload links, text, or screenshots and rate the severity of content. It uses a multi-model AI system, including RoBERTa, Twitter RoBERTa, DehateBERT, and the Vidgen Model, to analyze and score hate speech with high accuracy.
Impact:
By combining accessible UX with cutting-edge AI, Hate Shield empowers users to identify, evaluate, and act on harmful content online. The project was recognized for its practical application, social relevance, and innovative use of technology to counter hate speech at scale.