An analysis by SubmitHub examined more than one million tracks released in July using its SH Labs detection system. It classified 23.2 per cent as fully AI-generated and another 15.3 per cent as AI-generated audio later modified by humans.
Together those figures produce a striking total of 38.5 per cent, though the finding reflects one detection system rather than the entire industry.
Music has absorbed technological change before, as synthesisers, sampling and Auto-Tune all provoked arguments over authenticity before becoming normal. Generative AI presents a different problem, since it can replace much more of the creative process at once.
Listeners may not even know that replacement has occurred. The argument is therefore no longer whether AI belongs in music, since it already does.
Almost 40 Per Cent Now Involves AI
Another streaming platform tells a similarly dramatic story. Deezer said in June that fully AI-generated tracks had exceeded half of its daily deliveries.
That figure was up from 34 per cent in late 2025 and just 18 per cent the previous April. Tens of thousands of synthetic tracks can now enter a single catalogue every day.
Volume may become as important as quality. A musician once competed against other musicians for a limited amount of listener attention.
Generative systems allow almost unlimited songs to be produced cheaply and rapidly. That multiplies the amount of material competing for the same ears.
Listeners Cannot Reliably Tell
Audiences may be poorly equipped to identify the difference. A 9,000-person study across eight countries asked participants to distinguish fully AI-generated tracks from human-made music.
Some 97 per cent failed to identify all of them correctly, and 52 per cent said the uncertainty made them uncomfortable. Even musical expertise offered little protection, since listeners with years of conservatory training could not reliably tell the difference either.
That complicates the comforting assumption that mediocre machine music will simply expose itself. If audiences cannot consistently hear the distinction, musicians cannot rely on craftsmanship alone to make their origin obvious.
Listeners still appear to care about that origin. Eighty per cent of respondents wanted fully AI-generated music labelled, and 73 per cent wanted to know when streaming services recommended it.
Musicians Disagree on the Tool
Artists themselves are far from united. Some established producers treat AI as another technological instrument, much like innovations that once frightened the industry.
Used to manipulate sounds or accelerate technical work, it can extend a musician’s abilities rather than replace them. Others draw the line at systems capable of producing finished songs from short prompts alone.
Their concern is not merely artistic purity. Generative models have been trained using enormous quantities of existing music, raising questions over consent and compensation.
A German court recently ruled that Suno had infringed copyright by using protected compositions without authorisation, ordering it to disclose profits and pay damages. More than 1,800 artists are now backing similar legal actions elsewhere.
Tastemakers Face a New Problem
Curators, DJs, radio programmers and critics have traditionally filtered abundance for everyone else. Their value came partly from hearing more music than ordinary listeners could and flagging what deserved attention.
AI makes that filtering job considerably harder. Tastemakers must now judge not only whether a song sounds good, but how it was made and who stands behind it.
Fraud adds another complication. Deezer said up to 85 per cent of streams of fully AI-generated music detected in 2025 were fraudulent.
Much of that involved artificial streaming designed purely to extract royalty payments, and the company now strips such streams from its calculations. Australia went further this week, excluding fully AI-generated recordings from its official charts and music awards.
Human Origin Becomes Information
The arrival of convincing synthetic music does not have to mean audiences stop caring about musicians; it may produce the opposite effect instead. People already value live performances, physical records and personal connection despite instant access to almost every recording imaginable.
Knowing that somebody practised an instrument or spent months refining a track can be part of why music carries meaning.
AI can reproduce many musical characteristics without reproducing the life behind them. For background listening or disposable playlists, that distinction counts for very little.
For fans who follow careers and form real relationships with artists, it counts for considerably more. Labelling therefore offers a more sensible response than pretending synthetic music can simply be removed.
Listeners can enjoy AI-generated work if they choose, and musicians who use it as one tool among many need not be lumped in with automated content farms. Nearly 40 per cent is remarkable because it suggests the transition has already outpaced cultural habits.
The next struggle will concern less the sound than trust, knowing what people are listening to, who made it and how much human creativity remains behind the play button.
Keep up with Daily Euro Times for more updates
Read also:
Bubble Swells: Europe Insists on Labelling AI Content
EU Cybersecurity Rules Create a New Kind of Cyber Slop
Does Britain Need a Ministry for Artificial Intelligence?
