Spotify wants AI clones of podcast hosts to read ads
Spotify is reportedly developing AI tech that can clone podcast hosts' voices to generate ads without them ever stepping into a booth. The tool would let creators produce localized or multi-language sponsor reads with their permission. Spotify hasn't officially confirmed the feature, but admitted it's actively exploring generative audio.
Podcast host-read ads command premium rates for a simple reason: listeners actually trust the host. But what if the host never actually spoke those words?
According to The Ringer founder Bill Simmons, Spotify is working on an AI feature that can generate host-read ads using cloned voices—meaning creators would not need to sit in a booth to read ad copy anymore.
Speaking on a recent episode of The Bill Simmons Podcast, Simmons teased the upcoming capability. He noted that hosts would need to give explicit approval, but that it opens up massive possibilities, including hyper-local targeting for events in specific cities or even generating ads in foreign languages.
Simmons, whose company was acquired by Spotify in 2020, prefaced his leak by saying he did not think Spotify would get mad at him for sharing the detail.
Connect the dots to AI DJ
When reached for comment, Spotify played coy, saying advertising represents an interesting canvas for future exploration while declining to confirm or deny Simmons' claims. It did, however, point out that daily podcast episodes discussing AI surged 500% over the past month.
Even without an official confirmation, the writing has been on the wall.
Spotify acquired voice synthesis company Sonantic in 2022 and has paired its technology with OpenAI models. That combo already powers the app's AI DJ, voiced by cultural partnerships lead Xavier Jernigan. Jernigan spent hours recording lines with different cadences, intentional breaths, and natural slang like tunes or bangers to make the AI sound convincingly human.
With hundreds of employees dedicated to personalization and machine learning, scaling that tech to existing podcaster audio is the obvious next step.
Why it matters
We think Spotify is trying to solve the biggest bottleneck in podcast monetization: human scalability.
Host-read sponsorships work because of personal connection, but podcasters cannot realistically record dozens of localized variants for different markets. Cloned voices would let advertisers run dynamic, localized campaigns—think a local concert ad in Chicago vs. one in London—while creators collect passive ad checks without lifting a microphone.
That said, this strategy walks a very thin ethical tightrope. The entire value of a host-read ad rests on authenticity. If listeners realize the host has never touched the product and is just leasing out their vocal likeness to an algorithm, credibility could drop fast. Tech giants have stumbled here before: Google faced heavy backlash when Duplex booked restaurant reservations without disclosing it was an AI.
If Spotify pulls this off with clear disclosure, it turns podcasting into a true programmatic ad machine. If it fumbles the transparency, it risks eroding the intimacy that made podcasts valuable in the first place.
Sources (1)
- techcrunch.com(opens in a new tab)Originaltechcrunch.com
Topic trackerSpotify's generative AI expansion
- Spotify clones podcaster voices for instant translation
- Spotify plans to scale AI into podcast summaries and audio ads
- Spotify wants AI clones of podcast hosts to read adsThis story