Equitable AI/ML Solutions for Podcasting

This case study presents the collaboration between Sounder and Urban One to address brand safety and suitability in podcasting through AI/ML models. It outlines how these models aim to create equitable opportunities for Black creators by challenging misconceptions about their market viability. The study highlights that traditional keyword-based solutions often exclude a significant portion of content, with 92% of Urban One's episodes being removed from monetization opportunities. By employing a more nuanced AI/ML approach, only 10% of episodes are deemed unsuitable. The research emphasizes the importance of accurately transcribing Black English to prevent misclassification and marginalization of Black creators. Furthermore, it discusses the unique cultural nuances present in Black creator content and the implications for advertisers. The study concludes that Sounder's technology can enhance the inclusivity of diverse voices in podcasting, allowing brands to engage effectively with these communities while ensuring fair monetization opportunities.