AI Hosts Face A Cold Reception From Podcast Listeners.

The human connection, as demonstrated by the effectiveness of host-read ads, has been one of podcasting’s strengths. It’s little surprise then that listeners aren’t keen to the idea of having those creators being replaced by voices generated with artificial intelligence. New survey results from Edison Research finds that two in three weekly podcast consumers have a negative reaction to the prospects of their favorite podcast host replacing themselves with an AI-generated version.

Edison added Custom survey questions were added to Edison Podcast Metrics in Q1 2026 to gauge consumer reaction to AI and the data shows a third say they would feel deceived,” “upset” or “skeptical” — while 28% say they would consider it a betrayal of their relationship with the podcast host.

Only 27% of weekly podcast consumers selected any positive emotion. Edison says 14% of weekly podcast consumers say they are “impressed” by the usage of an AI-generated host. Roughly one in 10 people said they were either “optimistic” or “fascinated” by the use of AI.

Others are taking a wait-and-see approach. Of the weekly podcast consumers surveyed, 18% said they would feel “indifferent” to their favorite podcast host using an AI-generated version of themselves on their podcast.

The Edison data backs up other findings that also highlight the depth of the host-listener relationship. Nearly six in ten respondents (59%) in a Clutch survey released in July said they feel a personal connection with the podcast hosts they regularly hear. And that connection has a measurable impact on purchasing decisions. More than four in 10 consumers (42%) have purchased a product or service based on a podcast host’s recommendation.

For the time being, most listeners don’t need to worry about their podcast being hosted by an AI voice. A survey released by RSS.com this summer found that while 56% of podcasters say they either regularly use AI or are experimenting with it, it wasn’t in front of the microphone. The most common applications include generating transcripts (35%), writing show notes and descriptions (33%), brainstorming topics (32%), creating artwork (24%), editing (22%) and repurposing episodes into social content (21%).