More than half of US podcast listeners support advertisers using artificial intelligence in podcasting to make sure the ads they’re hearing are as relevant as possible, according to a new study by Acast.
The study was conducted by Acast Intelligence, which asked 400 weekly US podcast listeners using survey tool OnePulse on their attitude towards the use of AI in specific areas within podcasting.
“At Acast, we know that podcast audiences tend to be more aware of and interested in technology compared to most other mediums, so it’s no surprise that our research found listeners to be so interested in AI,” said Acast research and insights director Tommy Walters. “However, what is unexpected about this study is just how open listeners are to AI infiltrating nearly every aspect of their current listening experience and for the technology to mold new ways for them to engage with the medium.”
Other ways that listeners support brands and publishers using AI to enhance podcast advertising is through voice simulations with 56% of respondents agreeing that AI can be used by podcasters to simulate them reading ads instead of actually doing host-read ads themselves.
A number of companies have already started to experiment with using AI to deliver host-read ads using artificial voices, including audio AdTech company AdsWizz. The platform announced in its State of Audio Technology Report 2023 that it is developing “unique tools” under its Technology Innovation Lab which allow advertisers to create voiceovers using artificial intelligence.
Acast is also now planning to use AI to offer intelligent podcast recommendations within its advertising marketplace, which now allows brands and agencies to purchase host-read podcast advertising on a self-serve basis.
Further findings from the study revealed that over 80% of the respondents agreed that AI in podcasting can be used towards improving their listening experience by enhancing the sound quality and making it more accessible to listeners worldwide by translating captions to various languages.
More than 70% also said that AI could be leveraged to make podcast discoverability easier for listeners by suggesting specific podcast episodes that are relevant to the topic being discussed within the episode they’re currently listening to and to help them discover relevant podcasts.
AI can also be beneficial for podcasters in helping them understand their listeners with 77% of respondents agreeing that it can be helpful to generate reports on listener drop off rates which can inform the creators on what is and isn’t working within the podcast. It can also be used to pick up menial tasks such as generating show notes and summaries of episodes, generating highlights from the episode, creating podcast art for titles, and so on.
Although this was least popular amongst the respondents, 53% of listeners within the study also support the use of AI to create realistic fictional content such as interviews between celebrities or roundtables. This is sometimes known as ‘deepfaking’, which involves using artificial intelligence to create convincing audio and video that mimics a real person.
Cases of this have already previously been done in podcasting, with one example being an AI-generated interview between Joe Rogan and Steve Jobs which appeared in October 2022. Since then, a new channel on YouTube titled The Joe Rogan AI Experience has launched, which includes fake interviews between the famous podcaster and other well-known controversial figures such as Andrew Tate and Donald Trump.