AI Video Startup Pippa Pays Artists for Their Style

A new generative AI startup, Pippa, is attempting to address the contentious issue of artists' work being used without permission for model training. The company's core product involves generating short video clips, but its unique business model focuses on directly paying human artists. Every time a user generates content derived from an artist's style, that artist receives compensation from Pippa.

Pippa's co-founders believe this ethical approach will attract customers and prove that AI firms can operate responsibly. They aim to move beyond the "bloody history" of AI development, drawing parallels to the early days of music sharing with Napster. Artists receive a small fee per generated image ($0.005) and video ($0.003 per second), along with a portion of a 5 percent royalty pool funded by subscription revenue.

Launched in May, Pippa currently has around 800 paying subscribers and has established licensing agreements with four artists, with more in discussion. Artists undergo a vetting process to ensure ownership of their work. Despite efforts to foster trust, Pippa's models still utilize broader datasets, meaning some content may originate from art scraped without explicit consent, a challenge inherent to many "ethical" AI startups.

Visually, Pippa's output currently resembles that of other text-to-video services, often imitating popular aesthetics with mixed success. The company plans to integrate more sophisticated models, like ByteDance's Seedance 2.5, but its ability to truly stand out hinges on securing more in-house artists and shifting public perception of generative AI.