Spotify's AI-Powered Prompted Playlists Now Blend Podcasts Into Personalized Mixes for Premium Beta Users

Spotify's AI Playlist Tool Expands Into Podcasts — And It's More Significant Than It Sounds

Spotify has quietly made one of its more interesting AI features significantly more useful. Prompted Playlists, the platform's natural-language playlist generator that lets users describe what they want to hear rather than manually curating tracks, is now extending its reach into podcasts. It's a beta update, yes — but one that speaks to a larger strategic shift in how Spotify thinks about discovery, engagement, and the competitive pressure it faces from every direction.

The feature update, announced on April 7, is rolling out to beta testers across six English-speaking markets: the U.S., Canada, Ireland, the U.K., Australia, and New Zealand. Users in those regions can now write natural-language prompts that explicitly request podcast content — something like "create a playlist with the biggest entertainment news from the past few days, covering music, film, and fashion" — and let Spotify's algorithm do the sifting across what the company says is a catalog of 34 million podcasts discovered weekly on its platform.

From Music to Podcasts: Why This Matters Beyond the Feature Itself

On the surface, this looks like a minor expansion of an existing beta tool. Dig a little deeper, and it starts to look like Spotify testing the boundaries of what an AI-powered media assistant can actually do.

Prompted Playlists launched in New Zealand late in 2024 before reaching the U.S. and Canada in January 2025. The core premise was elegant: instead of being handed algorithmic suggestions based on opaque listening history signals, users could describe a mood, a genre, an artist vibe, or even an abstract concept, and the system would generate a playlist accordingly. It still incorporated listening history and what Spotify calls "vibes," but the user held the steering wheel. That combination — user intent layered over behavioral data — is a meaningfully different approach from pure algorithmic curation.

Extending that to podcasts isn't just a feature add. It represents Spotify treating its entire audio library — music, podcasts, and potentially audiobooks down the line — as a unified content pool that a single AI interface can navigate. That's a compelling proposition, and one that none of its major competitors have clearly articulated yet.

The Discovery Problem Podcasting Has Always Had

Podcast discovery is, charitably speaking, a mess. The medium has grown explosively over the past decade, but the tools for finding relevant shows have not kept pace. Search requires you to already know what you're looking for. Recommendation engines surface the same mega-hits repeatedly. Word of mouth remains one of the most effective discovery channels — which says a lot about how inadequate the algorithmic alternatives have been.

Spotify's own data point — 34 million podcast discoveries "every week" on the platform — is striking, but context matters. That figure likely reflects total instances of a user encountering a new show, not necessarily intentional searches. The friction of finding a genuinely relevant podcast on a topic you care about remains high, particularly for niche subjects or emerging shows that haven't yet accumulated the listener volume needed to surface prominently in standard recommendation feeds.

A natural-language prompt interface attacks this problem differently. Rather than asking a user to browse categories or guess at search terms, it invites them to express intent in plain language. "Find me science podcasts about space exploration that are accessible to non-experts" is a reasonable thing a person might want — and a reasonable thing an AI model should be able to act on. Whether Spotify's implementation delivers on that promise in practice will depend heavily on how well the system interprets varied prompt styles and how effectively it handles the long tail of the catalog beyond the obvious flagship shows.

Personalization With a Refresh Button

One underappreciated element of this feature is the refresh cadence control. Users can set their Prompted Playlist to update daily, weekly, or not at all. That flexibility matters more than it might initially seem.

Spotify's existing personalization tools — Discover Weekly, Release Radar, Your Top Mixes — operate on fixed schedules that users cannot control. Discover Weekly refreshes every Monday, full stop. If you discovered it on a Thursday and wanted fresh recommendations by Saturday, you were out of luck. Prompted Playlists breaks that rigidity, allowing users to define not just what they want but how often the list should evolve. For podcast listeners following ongoing news cycles or episodic content, daily refresh could be genuinely useful. For music listeners building a specific mood playlist, the static option preserves the curation effort.

This kind of granular control reflects a growing understanding that "personalization" isn't one-size-fits-all — different content types have different consumption rhythms, and the tools serving them should account for that.

The Competitive Stakes

Spotify isn't developing these AI features in a vacuum. Apple Podcasts, Amazon Music, YouTube Music, and a range of independent podcast apps are all competing for ears and engagement time. Apple in particular has invested heavily in podcast infrastructure since its acquisition of Anchor's competitor landscape reshaped the market. YouTube's combination of video and audio podcast content gives it a discovery advantage through its existing recommendation engine — one of the most powerful in media.

What Spotify is betting on is that a unified audio experience — music, podcasts, audiobooks, all accessible through a single conversational interface — becomes stickier than any individual content category alone. If a user can express "I want to start my morning commute with a news podcast, then transition to focus music for work," and have Spotify build and maintain that queue automatically, the platform becomes infrastructure rather than just an app. That's a fundamentally different relationship with the user, and a much harder one for competitors to replicate quickly.

What Beta Users Should Actually Expect

It's worth tempering expectations appropriately. This is a beta feature, still limited to specific markets and a subset of users within those markets. Early implementations of AI-driven discovery tools frequently struggle with prompt ambiguity, catalog gaps, and the inherent challenge of translating subjective language into reliable content matches. "Chill vibes for a Sunday afternoon" is easy. "Podcasts about the intersection of behavioral economics and urban planning" is considerably harder.

Spotify itself has acknowledged the feature could evolve — including potentially introducing usage limits — which suggests the company is still calibrating what sustainable, scalable deployment looks like. The introduction of podcast support now is likely partly about gathering data on how users actually phrase podcast-related prompts, which will be essential for improving accuracy over time.

For users in the eligible beta markets, the entry point remains straightforward: tap "Create," select Prompted Playlists, and describe what you want. The addition of podcasts to that prompt vocabulary is live as of this week. The more interesting question is what Spotify adds to that vocabulary next — and whether the feature graduates from beta before the competitive window it's trying to exploit starts to close.