‘Verified by Spotify’ Badges Expand to Podcasts Amid AI Slop Wave

Photo Credit: Flipsnack

As AI uploads continue to inundate DSPs, Spotify has expanded its verification badges to the podcasting side, where programs are being reviewed “against Spotify’s standards for authenticity and trust.”

The streaming service officially brought “Verified by Spotify” to podcasts today, about three weeks after unveiling the verification measure for artist pages. Most are aware that both moves arrived amid an AI slop avalanche (and no shortage of related criticism from fans as well as real artists) on music platforms.

Evidently, though, machine-generated audio is also prevalent in podcasting. Enter the launch of Verified by Spotify for podcasts, which, assuming they meet the service’s review standards, will be identified with a namesake badge featuring a green checkmark.

Time will tell what those standards look like in practice; “select shows” are receiving the label out of the gate, with plans in place for a wider rollout “over the coming months.”

But at the top level, Spotify intends to verify programs it can “confidently authenticate” based on “sustained listener activity,” compliance with its rules, and “audience authenticity, including safeguards against fraudulent or bot-driven listenership.”

Meanwhile, the service has formally expanded its ban on unauthorized AI deepfakes to podcasts. “Spotify will remove podcast shows and content that impersonate another creator or host’s likeness without permission, whether that’s using AI voice cloning or any other method,” per the announcement.

The implementation of verification measures is a long-overdue step in the right direction for Spotify, which, to put it generously, has for years had relaxed upload-moderation policies.

Some will recall the on-platform presence of drug-plugging “podcasts” last year, and as things stand, one need only search for vague terms like “funny” to find plenty of content (such as “Funny sounds/ringtones”) classified as podcasts.

An obvious-but-important point: A significant portion of these uploads could have been (and can still be) decommissioned with simple filters based on title/description keywords, length (a podcast episode probably shouldn’t run less than 30 seconds), and more.

However, Spotify’s been pulling out all the stops (including by embracing uploads en masse) to maximize user retention as it locks horns with competitors such as YouTube in a battle for podcasting supremacy.

Even so, Verified by Spotify’s active buildout raises an interesting question: With proper music and podcasts now featuring unique identifiers, presumably part of a plan to help users avoid AI slop altogether in the long term, who’s the AI slop for?

Scammers come to mind here. According to Deezer, the overwhelming majority of streams on machine-generated tracks are fraudulent, and the uploads are hitting DSPs at a faster rate than ever.

Back to Spotify, it’s worth reiterating the ongoing internal development of music-generation tools. Will said tools ultimately be set loose on a mountain of AI garbage, thereby enabling the unlimited one-stop “creation” of derivative tracks?