EVALUATING THE EFFECTIVENESS OF AI-POWERED CONTENT MODERATION ON INSTAGRAM: A CASE STUDY OF HATE SPEECH DETECTION IN 2024

Abstract

With the exponential growth of user-generated content on social media platforms, effective content moderation has become paramount to maintaining safe and inclusive online environments. Instagram, as one of the world’s largest social media platforms, relies heavily on artificial intelligence (AI) to detect and mitigate harmful content, including hate speech. This study evaluates the effectiveness of Instagram’s AI-powered content moderation system in detecting hate speech in 2024. Drawing on recent technological developments, platform disclosures, and academic critiques, the paper analyzes Instagram’s AI algorithms’ strengths and limitations in hate speech identification, the role of human moderators, and the challenges posed by contextual nuances and cultural diversity. The findings reveal that while AI significantly enhances scalability and speed in content moderation, it continues to struggle with understanding context, irony, and local linguistic variations, leading to both false positives and false negatives. The study underscores the necessity of a hybrid moderation model combining AI efficiency with human judgment to optimize hate speech detection and uphold freedom of expression.