Streaming is drowning in old songs: two-thirds of plays are music older than five years

The key facts

About 60% of streams in the UK and US in the first half of 2026 were songs older than five years (Luminate).
18 of the 20 best-selling songs of the year in the UK were released more than a year ago; Zara Larsson's 2015 hit Lush Life was the eighth best-selling.
Spotify removed 75 million “spam tracks,” but on Deezer AI accounts for only 0.5% of listens.
Spotify removed 75 million 'spam tracks', but on Deezer AI accounts for only 0.5% of plays.

Nostalgia beats new releases

More music is being created than ever before: over 100,000 songs are uploaded to streaming services every day. However, according to the analytics company Luminate, which has tracked music consumption since 2008, in the first half of 2026 about 60% of listens in the UK and US went to tracks released five or more years ago. Streaming of current music — released less than 18 months ago — lags noticeably behind the catalogue of older recordings.

Charts full of old hits

In the UK, 18 of the 20 best-selling songs of 2026 were released 12 months ago or earlier. The eighth-biggest hit of the year is Zara Larsson's Lush Life, released in June 2015; Michael Jackson's Billie Jean (1983) sits at number 22, and Fleetwood Mac's Dreams (1977) at number 30. A similar picture holds in the US: in the two most-streamed genres, R&B and hip-hop, about 60% of streams are old hits. Even when new songs break through to the top, they hold the lead for weeks: the Sam Fender and Olivia Dean duet Rein Me In set a UK chart record with 23 non-consecutive weeks at number one, while Ella Langley's Choosin' Texas became the longest-running number one in the history of the American chart.

Concentration and pressure on new artists

According to the UK's Competition and Markets Authority, almost two-thirds of streams go to music by “just 0.4% of the most successful artists”, and to reach the top of the charts, hits need millions of plays per day. J Erving, founder of the distribution company Human Re Sources behind Raye, in a recent interview with Billboard noted that today “it's easier for people to be passive fans,” and wondered whether listeners are truly deeply engaged with artists rather than just individual songs. Apple Music chief Oliver Schusser told the BBC that the service is working to promote “artists of tomorrow” in its curation, since future headliners like Super Bowl halftime show performers “need somewhere to start.” Sofar Sounds, an organisation staging intimate pop-up concerts, held the Sofar Fringe festival in London's Shoreditch in October, featuring, among others, rapper J Casa, and is launching the Human Discovery Movement campaign: listeners can text a “mystery number” and receive voice recommendations of new music from curators.

AI playlists and suspicious “artists”

Spotify, Apple Music and Amazon Music are promoting AI curation, allowing subscribers to generate playlists from a text prompt. Musicologist Anna-Kaisa Kaila believes that “the potential for bias absolutely exists”: if the training data contains few tracks of a certain, commercially non-dominant style, the system will find it harder to surface them. London-based independent artist Chiminyo claims that in a “relaxing jazz piano playlist” with more than 370,000 saves on Spotify, the top performers with hundreds of thousands of monthly listeners have unverified accounts about which almost nothing can be found online, and the tracks sound almost identical — in his view, they are synthetically mass-produced. Spotify responded that it has introduced platform rules against impersonation and is working to improve filtering of AI-related issues, while acknowledging that the use of AI in music production lies on a spectrum. At the same time, despite the flood of AI music — Spotify removed 75 million “spam tracks” — on Deezer it accounts for only 0.5% of listens.

Why it matters

The situation directly affects emerging musicians whose income depends on streaming: with listening concentrated on the catalogue and the top 0.4% of artists, it becomes harder for new performers to break through, and algorithmic playlists, according to the experts surveyed, tend to recommend what is already familiar. For listeners, the practical consequence is the growth of services and initiatives with human curation — from Apple Music to Sofar Sounds — while platforms face the need to tighten filtering of AI content. It remains unclear whether the balance between old and new music will change: the available data covers only the first half of 2026.