Lorde is pushing back against Spotify’s latest AI-powered feature. The singer criticized the streaming platform after discovering an inaccurate AI-generated explanation attached to one of her songs, arguing that it undermines artistic interpretation.
In a post shared on Instagram, Lorde addressed Spotify directly, writing:
“Hey @spotify i’m gonna go out on a limb n say we don’t want this. Not only is this inaccurate (not the song i did that in) but reducing a song to an ai generated meaning right at the source feels like it limits free interpretation imo. At least make it possible for artists to opt out please.”
Her comments were directed at Spotify’s new “About the song” beta feature, which uses artificial intelligence to generate brief descriptions explaining songs and their background. In Lorde’s case, the summary attached to “Current Affairs” incorrectly referenced a live performance in which she stripped down to her underwear while a dancer poured water over her during the Ultrasound World Tour. According to the singer, that moment belonged to an entirely different song, making the AI-generated explanation factually incorrect.
Beyond the mistake itself, Lorde’s criticism focused on a broader issue: whether AI should be defining the meaning of music at all. She argued that automatically presenting a single interpretation directly within Spotify limits the freedom listeners have to form their own emotional connections with a song. Rather than calling for the feature’s removal, she urged Spotify to allow artists to opt out if they do not want AI-generated descriptions accompanying their work.
The controversy arrives as Spotify continues expanding its AI initiatives. Earlier this week, the company introduced Talk to Spotify, a conversational AI assistant that lets eligible Premium subscribers search for music, podcasts and audiobooks through voice or text. The feature, currently available in beta for Premium users in the United States, Ireland and Sweden, can answer questions about albums, recommend songs based on mood, and even explore users’ listening history.
Leave a Reply