Here’s what’s on our plate today:

• 🎶 AI is remixing the music industry—will artists and labels keep up?

• 🧐 Vote if you love, hate, or fear robot bangers.

• 🛠️ Fresh tools to turn your wildest melodies into full songs.

• ⚡ Can detection tools really keep up with a tidal wave of synthetic tracks?

Let’s dive in. No floaties needed…

The Laboratory

How AI is reshaping music, rights, and creative control

Whether it’s the onset of the monsoon or a long drive down a country road, music has the power to encapsulate the beauty of both. And why shouldn’t it be! Since prehistoric times, humans have crafted flutes from bones, later evolving to string instruments, radio, digital music players, and now, generative AI. Regardless of the era, humans have always had not just a strong affinity but a need for music.

Whatever the instrument or the technology used in the production of music, the involvement of humans has been central throughout it all. Musicians now face competition from artificial intelligence models that can not only produce music at scale but also mimic popular artists.

Prominent tools in AI music generation include Google’s Magenta, Endless, Boomy, and Voicemod AI Music. OpenAI is also working on its foundational project in AI music generation, called Jukebox. Although not yet commercially available, it illustrates how tech companies envision the future of music creation.

As companies race to develop increasingly sophisticated music-generation tools, artists worldwide are scrambling to find a way to track synthetic content, label it, while trying to ensure that their existing work does not become fodder for training under-development and future models.

The AI duet that shook the music industry

The music industry’s nightmare came true in 2023, when “Heart on My Sleeve,” a convincingly fake duet between Drake and The Weeknd, racked up millions of streams.

The song, created by a TikTok user going by the alias Ghostwriter977,  was AI-generated and used voice cloning technology to mimic Drake and The Weeknd. The vocals were convincing enough to lead listeners to believe it was a genuine collaboration. The song accumulated approximately 600,000 streams on Spotify, 275,000 views on YouTube, and 15 million views on TikTok.

At the time, Universal Music Group (UMG), which represents Drake and The Weeknd, issued takedown notices, resulting in its removal from platforms. However, the damage was already done.

The song exposed cracks in how the industry controlled synthetic content and shifted the focus from trying to stop AI-generated music to tracing its origins and managing it in a way that would stop synthetic content from drowning out human artists.

While the industry was figuring out how it wanted to tackle the challenge, artists were divided on how they saw the advent of AI tools.

The double-edged sword of AI in music

While the debate about the pros and cons of AI in the music industry raged on, no one could deny the role technology played in lowering the entry barrier. With AI, anyone with a text prompt can create songs, and tools assist with mixing, mastering, or voice isolation.

This, though, creates a new set of problems. The lower entry bar ushered in by generative AI relies on models that learn from massive datasets, often scraped without permission, from around the world.

This not only raises ethical concerns about bias, representation, and unlicensed use but also threatens those who rely on musical income—like stock-music creators or jingle writers—as machines can produce “acceptable” tracks en masse. As a result, musicians and publishers resorted to filing lawsuits against AI companies like Meta Platforms, Microsoft-backed OpenAI, and Anthropic, all of whom have been accused of using licensed data to train their models.

While the industry is looking to get its fair share, not all musicians view AI as a threat. Artists like Holly Herndon, Arca, YACHT, and Brian Eno embrace AI as a collaborator, using it for creative exploration rather than full autonomy. These artists even argue that AI’s tendency to produce “hallucinations” and other unpredictable or incoherent outputs, though dangerous in other contexts, is a means of inspiration.

Live coding performances and tools that generate stems are examples of this hybrid process.

Takedowns don’t work, licensing might

Once over the initial shock, the music industry started working on a new category of infrastructure. One that’s built not to stop generative music outright, but to make it traceable.

The modus operandi: to embed detection systems across the entire music pipeline, including in tools that are used to train models, platforms that host music, and licensing databases.

The objective is twofold: to trace AI-generated music and to ensure it is tagged and controlled throughout the music pipeline.

The goal is being worked on in tandem with startups that are building detection into licensing workflows. Platforms like YouTube and Deezer have developed internal systems to flag AI-generated audio as it’s uploaded and shape how it surfaces in search and recommendations. Other music companies, including Audible Magic, Pex, Rightsify, and SoundCloud, are also expanding detection, moderation, and attribution features across everything from training datasets to distribution.

Some companies are going to the extent of analyzing what goes into an AI model to determine how much of the tracks generated by it are borrowed from specific artists or songs.

Overall, the industry aims to ensure artists have more precise licensing controls, with royalties based on creative influence rather than post-release disputes.

Platforms walk the tightrope between profit and protection

While musicians and publishers are prioritizing early detection of synthetic content, platforms like Spotify are open to AI-generated music unless it blatantly infringes copyright. Labels like Universal Music Group are also investing in AI but lobbying for fair frameworks via initiatives like the Human Artistry Campaign.

It is not difficult to assess why platforms may be favourable to AI-generated music. With over 100,000 tracks uploaded daily, platforms have economic incentives to favor royalty-free AI content, potentially diverting royalties from human creators.

As AI-generated music proliferates, now constituting around 18% of daily uploads on Deezer, or over 20,000 tracks a day, the industry has turned to tracking rather than banning synthetic content.

Deezer has begun labeling AI-generated tracks, removing them from editorial playlists, and excluding fraudulent streams from royalties. While AI-generated music makes up only 0.5% of total streams, as much as 70% of it is flagged as fraudulent, often produced by streaming farms.

A surge in lawsuits, from Sony, UMG, and Warner against AI creators like Suno and Udio, highlights the growing legal backlash over dataset use and copyright infringement  . While platforms like YouTube, Audible Magic, and SoundCloud are developing proprietary classifiers, startups such as Vermillio are using stem-level metadata tagging for pre‑release clearance  .

This infrastructure shift, from reactive takedowns to proactive control, signals a pivotal moment. As the music world adapts, artists and platforms must balance innovation with transparency, ensuring that AI becomes a tool for artistic enhancement rather than exploitation.

3 Things Worth Trying

• AIMusic - AI transforms melodies or lyrics into full tracks. Effortless songwriting and inspiration. Explore and download creations easily.

• Suno AI - Generate full songs with vocals from text prompts.

• Soundraw - Custom AI music by mood, genre, and instruments.

Thursday Poll

Rate this edition