This research introduces an innovative AI framework that leverages various large language models to identify and mitigate bias in news articles, addressing a critical challenge in information integrity.

• Systematically evaluates six different LLMs (GPT-4o, GPT-4o Mini, Gemini Pro, Gemini Flash, Llama 8B, and Llama 3B) for bias detection performance
• Provides an automated pipeline to neutralize biased content while preserving factual information
• Demonstrates significant improvement over traditional human moderation in terms of scalability and objectivity
• Particularly effective for security-related news coverage of crime and societal issues

For security professionals, this research offers a practical tool to ensure more balanced information consumption and distribution, potentially reducing polarization in security discourse and improving public trust in reporting.

Neutralizing the Narrative: AI-Powered Debiasing of Online News Articles