Curiosity Stream, the science-focused streaming service that has long struggled to attract subscribers, has taken a bold step into a new era of business by leveraging its vast library of content in a lucrative way: licensing it to AI companies. The platform, launched in 2015 by Discovery Channel founder John Hendricks, initially focused on attracting viewers with high-quality science and educational programming.
While its subscription-based model has yet to generate significant revenue, Curiosity Stream's original content is in high demand among large language models (LLMs) and other AI systems. In fact, the company expects licensing deals with AI companies to surpass its existing subscription business by 2027.
The shift towards AI licensing is a strategic move that could significantly boost Curiosity Stream's bottom line. According to CEO Clint Stinchcomb, the platform is now licensing over 300,000 hours of its own content, as well as third-party content, to support the growth of LLMs and other AI systems.
This new revenue stream has proven highly successful, with revenue from IP licensing reaching $23.4 million through September this year โ more than half of what Curiosity Stream earned from subscriptions in 2024. Stinchcomb believes that the opportunity for licensing will only grow as more open-source models become accessible, potentially attracting hundreds or even thousands of companies seeking to fine-tune their AI models with high-quality video content.
While some have questioned whether this strategy is sustainable, particularly given ongoing litigation and the uncertain nature of the AI market, Curiosity Stream's executives remain optimistic about its long-term prospects. As Stinchcomb noted, the platform aims to build "long-term relationships" with IP holders and explore a range of licensing agreements beyond traditional training rights.
As other content-centric companies consider new revenue streams to stay competitive, Curiosity Stream's success in AI licensing serves as a reminder that even established players can adapt and thrive in rapidly evolving markets. For now, at least, the platform is poised to ride the wave of AI-driven demand for high-quality video content.
While its subscription-based model has yet to generate significant revenue, Curiosity Stream's original content is in high demand among large language models (LLMs) and other AI systems. In fact, the company expects licensing deals with AI companies to surpass its existing subscription business by 2027.
The shift towards AI licensing is a strategic move that could significantly boost Curiosity Stream's bottom line. According to CEO Clint Stinchcomb, the platform is now licensing over 300,000 hours of its own content, as well as third-party content, to support the growth of LLMs and other AI systems.
This new revenue stream has proven highly successful, with revenue from IP licensing reaching $23.4 million through September this year โ more than half of what Curiosity Stream earned from subscriptions in 2024. Stinchcomb believes that the opportunity for licensing will only grow as more open-source models become accessible, potentially attracting hundreds or even thousands of companies seeking to fine-tune their AI models with high-quality video content.
While some have questioned whether this strategy is sustainable, particularly given ongoing litigation and the uncertain nature of the AI market, Curiosity Stream's executives remain optimistic about its long-term prospects. As Stinchcomb noted, the platform aims to build "long-term relationships" with IP holders and explore a range of licensing agreements beyond traditional training rights.
As other content-centric companies consider new revenue streams to stay competitive, Curiosity Stream's success in AI licensing serves as a reminder that even established players can adapt and thrive in rapidly evolving markets. For now, at least, the platform is poised to ride the wave of AI-driven demand for high-quality video content.