The Audio Landscape is Overrun by AI ‘Podslop’ Too

Photo Credit: Austin Distel

The rise of AI isn’t just leading to music slop on streaming services, but a serious increase in automated podcast content flooding the landscape.

It’s not exclusive to the music industry. The sheer volume of AI-generated podcast content is beginning to affect traditional discovery methods across the industry that podcast creators and their listeners rely on. Over a period of just nine days, nearly 39% of new podcast feeds were identified as potentially AI-generated.

This rising trend in “podslop” was recently illustrated by data from the Podcast Index, an open-source tracking platform, and subsequently reported by Bloomberg. Most of these shows target high-volume search terms, such as health and wellness or celebrity biographies.

Some platforms, such as Apple Podcasts, require creators to disclose if a significant portion of their show was created using AI. Others, like Spreaker, have started manually labeling content as AI-generated—but the speed at which such content is produced has quickly outpaced human-led moderation to label it.

Unlike AI-generated music, which runs the risk of copyright infringement, AI-generated podcasts are much easier to create, and there’s money to be made in doing it “well.” The rapid creation of AI-generated podcasts covering a broad range of content is designed to catch search traffic and the programmatic ad revenue that comes with it, all with minimal human involvement.

Networks using AI tools to generate their content can output thousands of episodes per week for as little as a dollar per episode. Some networks have already built up a portfolio of over 4,000 shows. Such a workflow allows networks to monetize through sheer scale, something with which human-created podcasts simply can’t compete. And that’s despite the fact that AI-generated podcasts are typically far lower quality than most human-hosted alternatives.

The desire to speed up and simplify production means there could certainly be plenty of use cases for AI in the realm of podcast creation. For example, Adam Levy described how he found success with an AI-generated podcast based on the Epstein files, and aims to produce 120 episodes a month—a tremendous undertaking for one person without the use of AI to synthesize massive amounts of data.

“I figured out how to do that same work in hours instead of months, and I am publishing everything I find, sourced and cited and verifiable, as fast as I can produce it,” Levy said.

But even major companies are getting involved. Amazon has created an AI “podcast” about products users search for when using the platform. The somewhat surreal feature tells you detailed information about the product and will even take questions from users live, like QVC or HSN.