Majors want to ban “AI slop” from the charts

Music charts are becoming the new front in AI regulation

Major record labels are no longer content merely to negotiate with music artificial intelligence companies: they are now looking at the place AI-generated tracks can occupy in the charts. According to a report published by The Verge, the three majors—Universal Music Group, Sony Music and Warner Music Group—are discussing common rules intended to prevent tracks produced entirely by artificial intelligence from entering the charts.

The issue is highly strategic. Charts are not merely editorial rankings or promotional tools. They have become indicators of value for the entire industry: they direct platform attention, feed marketing campaigns, influence radio programming, contribute to building a career and serve as a benchmark for commercial partners. In this context, the proliferation of automated tracks, produced quickly and potentially released in very large numbers, is seen by labels as a risk to artists' visibility, but also to the very credibility of rankings.

The Verge reports that the discussions concern shared eligibility conditions. The main target would be what the industry sometimes calls “AI slop,” a pejorative term for content generated on a massive scale, with little human creative intervention and without necessarily meeting artistic, technical or documentation standards comparable to those expected of a conventional music release.

The term should be handled with caution. It is neither an established legal category nor a sufficiently precise technical description to distinguish, on its own, a software-assisted human work from a track created from start to finish by a generative system. Its use nevertheless reflects a very concrete concern: music can be made, adapted, distributed and tested on a scale that traditional creation and promotion cycles cannot achieve.

The proposal attributed to the majors therefore marks an important development in the debate. Until now, public discussions about music AI have focused primarily on copyright, model training, vocal imitations, artists' consent and compensation. The question of charts introduces another dimension: even when content is distributed, and even when it is technically listened to by real users, should it nevertheless be treated as an equivalent track in commercial competition?

This distinction is sensitive. Excluding a track from the charts does not necessarily mean banning it from a platform. Rather, it means separating market access from access to a form of institutional recognition. The majors thus appear to want to act not only on the production or licensing of content, but also on the mechanisms that turn a listen into visibility, and visibility into economic value.

After licensing, control over commercial recognition

The initiative comes at a time when relations between the music industry and AI players are being reshaped. The majors have long adopted a primarily defensive position toward music generators: their catalogs, their artists' recordings and song lyrics represent essential assets, and their use in the training or operation of AI models raises legal and compensation issues that remain central.

But the landscape can no longer be reduced to a simple opposition between rights holders and technology companies. The licensing agreements mentioned in the editorial brief show that the majors are also considering contractual relationships with certain music AI players. This development does not mean that tensions have disappeared. Rather, it indicates that the groups are seeking to define the conditions under which AI can become a commercially usable tool, without dissolving the rights, revenues and editorial control that structure their business.

Discussions about charts fit into this logic. Licensing a technology or negotiating the use of a catalog answers a first question: under what conditions can a company use works, recordings or voices? Defining criteria for charts answers a second: what commercial and cultural place should the results produced by this technology receive?

The two issues are linked, without being identical. An AI tool may, for example, be used as part of a human production, with songwriters, performers, a producer and a team that retain a substantial role. Conversely, a catalog of entirely generated songs may be released at scale with little information about its origin, production method or the real identity of those responsible. A chart policy must therefore, if it is to be credible, avoid confusing technological assistance with full automation.

This is precisely where transparency emerges as a key issue. The points reported by The Verge suggest that the contemplated framework could lead to more information about the use of AI in music releases. To decide whether a track is eligible, it is necessary to know how it was designed, what share is attributable to human beings, who controls the associated rights and whether the data declared by the distributor are sufficiently reliable.

Metadata, often invisible to the public, then take on a political dimension. They determine the identity of a recording, its authors and rights holders; they also determine the proper allocation of payments. In an environment where the same infrastructure can produce a very large number of tracks, incomplete or misleading information can quickly contaminate recommendation mechanisms, fraud-detection tools and listen-counting systems.

The majors do not control the charts on their own. Streaming platforms, distributors, organizations that compile rankings and specialist media all play a role. This is why the search for “common rules” is significant: an isolated decision by one label would have limited scope in the face of millions of independent releases and the diversity of distribution channels. Shared criteria, on the other hand, could create a market standard, even in the absence of a specific law devoted to generative music.

For music groups, the objective also appears to be preserving a difference between a track's technical availability and its recognition by the industry. A platform may host a work, a distributor may deliver it, but a chart may decide that it does not meet the conditions required to compete with productions led by human artists. This distinction could become a compromise between a ban that is difficult to implement and unlimited acceptance of automated production.

Why “AI slop” worries the platform economy

The majors' concern is not based solely on a cultural opposition between human creation and machine. It also concerns the very concrete economics of music platforms. In streaming, attention is scarce. Every position in a chart, every addition to a playlist and every algorithmic recommendation can generate considerable leverage. When a growing volume of tracks competes for these spaces, competition is no longer limited to quality or notoriety: it also becomes a matter of publishing capacity.

AI systems sharply lower barriers to production. This decrease can have positive effects: it can help musicians experiment, compose demos, find arrangement ideas or overcome certain technical limitations. But it also makes possible an industrial stream of tracks designed to capture listens, exploit keywords, imitate popular genres or occupy functional niches. Ambient music, tracks intended for sleep, concentration or workout playlists, variations on highly codified genres: all are segments where quantity can become a competitive advantage.

In this model, the expression “AI slop” does not refer solely to an aesthetic assessment. It denotes the risk of production optimized for metrics rather than for a lasting relationship between an artist and their audience. If fully generated tracks manage to enter the charts thanks to large streaming volumes or aggressive distribution practices, they can divert attention that would otherwise have benefited identifiable artists, songwriters, performers and producers.

The problem is all the more delicate because charts are supposed to reflect real usage. Excluding AI-generated tracks raises a question of principle: should a chart measure, without filtering, what is consumed the most, or should it also preserve a definition of what it considers an eligible musical work? Historically, chart rules have never been entirely neutral. They set release windows, availability criteria, counting methods and rules against certain forms of manipulation. AI adds a new variable to these trade-offs.

The difficulty will be drafting criteria that do not rely on an imprecise intuition. The phrase “entirely generated by AI,” mentioned in the brief, seems clearer than the word “slop,” but it immediately raises questions. What does entirely mean? Does a song whose instrumental is generated but whose lyrics were written by a person fall into this category? What about a synthetic voice directed by a human performer? And how should tracks be treated when part of their production has been assisted by automated tools already common in studios?

The boundary between tool and author is far from obvious. Pitch-correction software, virtual instruments, sound libraries, stem-separation features and audio-processing tools have gradually transformed music production for decades. An effective rule will probably not be able simply to detect the presence of AI. It will have to focus on the degree of the system's autonomy, the creative and economic role of identifiable people, and the quality of the information provided at the time of distribution.

The majors have an interest in this precision. A policy that is too broad could affect their own artists and producers, who already use assistive software in their creative processes. A policy that is too narrow, conversely, would not make it possible to limit automatically produced catalogs. The discussion is therefore not about a simplistic opposition between innovation and tradition; it is about defining a framework in which innovation does not turn the charts into a privileged playground for the most automated publishers.

Another issue concerns public trust. A chart retains its value because listeners, media and industry partners grant it a form of credibility. If users feel that rankings are filled with opaque tracks, artificial identities or content produced to manipulate discovery mechanisms, the chart brand itself may be weakened. For the majors, protecting access to the charts therefore also means protecting a market signal.

Common rules that are difficult to enforce, but potentially structuring

The desire to build a common framework does not guarantee its implementation. The majors can influence distribution and industry negotiations, but they do not alone represent all music published on platforms. Independent labels, self-produced artists, digital aggregators and technology companies will also have to be involved if the rules are to have a real effect on the charts.

The question of verification will be decisive. A voluntary declaration of AI use may provide a first level of transparency, but it depends on the good faith of the entity delivering the track. Technical checks could complement this system, without providing a perfect solution: detecting generated content is complex, especially when human and synthetic elements are mixed or when an audio file has been altered after generation.

The risk of false positives is real. A track recorded by musicians and processed with modern tools should not be equated with an entirely automated creation solely on the basis of sonic characteristics. The risk of false negatives also exists: generated content can be reworked or presented in a way that conceals its origin. This is why metadata, disclosure obligations and distributor accountability mechanisms will probably be as important as detection tools.

This approach is a reminder that AI regulation does not take place solely through courts or parliaments. In cultural industries, technical and contractual standards can sometimes have an immediate scope. A platform can require certain information before publication. A distributor can impose guarantees. A charting organization can set eligibility criteria. These decisions do not replace the law, but they concretely determine the behavior of players who depend on these infrastructures.

For artists, the outcome could be ambivalent. On the one hand, rules limiting entirely AI-generated tracks in the charts could reduce the risk of being drowned out by volumes of automated content. They could also encourage better recognition of creators' identity and the provenance of works. On the other hand, poorly designed criteria could penalize independent artists who legitimately use AI tools due to a lack of access to large production budgets.

The distinction between creative use of AI and opportunistic industrialization of publishing must therefore remain central. In electronic music, rap, pop or experimental music, the adoption of new technologies is part of the history of artistic practices. The sampler, sequencer, computer and autotune all sparked debates over authenticity before becoming integrated, to varying degrees, into studio practices. Generative AI nevertheless differs through its ability to produce not only a sound or an effect, but a complete song, sometimes with lyrics, structure, voice and an associated visual identity.

The historical comparison has its limits. Previous tools did not all raise the question of training on massive corpora of protected works, nor that of convincing imitation of existing performers. Nor did they make it possible, with the same ease, to multiply releases under fictitious artist names. The rules contemplated by the majors therefore do not simply seek to arbitrate a new studio technique; they respond to a potential transformation of the supply itself.

The move could also shift competition between services. Platforms that highlight clear information about the origin of tracks, the people involved and the AI status of content will be able to differentiate themselves. Conversely, services that allow opaque catalogs to thrive could face criticism from rights holders, artists and the public. Transparency is no longer merely a compliance issue: it is becoming an element of trust and reputation.

What this chart battle could mean for France and Europe

For the French-speaking market, the debate goes far beyond the interests of the three American or international groups. France has a dense network of independent labels, publishers, producers, studios and artists for whom digital discoverability is essential. Access to playlists and charts can carry disproportionate weight for projects that do not benefit from marketing power comparable to that of major global releases.

In this environment, a rapid rise in entirely generated content could affect emerging artists in particular. Streaming revenue is already distributed in a universe where the abundance of releases makes it difficult to emerge. If automated catalogs manage to occupy recommendation spaces or chart positions, pressure on human artists could increase, especially in genres where consumption relies heavily on playlists.

A possible common rule developed around the majors would also have indirect effects in Europe, even if its precise arrangements are not yet public. Standards imposed by major catalog holders, global platforms and distributors often tend to spread beyond their market of origin. For French players, the challenge will be not to be subjected to criteria designed elsewhere without taking part in the debate over definitions, disclosure obligations and avenues of appeal.

The European dimension is particularly important because the continent has made the transparency of AI systems a major regulatory focus. But music presents its own difficulty: it is not enough to know that an AI tool was used. It is also necessary to understand what that information implies for related rights, copyright, compensation, the performer's identity and eligibility for visibility mechanisms. Simple labeling could be useful, without by itself resolving the economic issues.

Professional organizations and collective management societies will therefore have a role to play in the emergence of understandable standards. They will have to defend the traceability of works while avoiding a definition of AI that would block ordinary creative uses. For a songwriter, an electronic music producer or a sound engineer, the ability to use new tools can be an opportunity. For an artist whose voice, style or repertoire is reproduced without authorization, it can become a direct threat.

The proposal mentioned by The Verge could thus herald a more explicit separation between several categories of music: works created by people with technological assistance; hybrid works, in which the generative share is substantial but framed; and entirely generated content, produced without significant human creative contribution. This typology is not yet an established rule, but it corresponds to the need to distinguish realities that the single label “AI” tends to conflate.

The next step will not only be to find out whether the majors reach agreement among themselves. It will be necessary to observe whether charting organizations, platforms and distributors adopt compatible definitions, and whether they publish procedures sufficiently clear for artists. A rule that excludes without explanation can create new forms of opacity; a documented rule, accompanied by verifiable declarations and avenues for challenge, can instead become a tool of trust.

In the long term, the value of charts could depend less on their ability to count every listen than on their ability to explain what they recognize. The rise of AI forces the industry to choose: to consider rankings as simple volume tables, or as cultural institutions that grant particular visibility to identifiable works and accountable creators. By seeking to exclude entirely generated tracks from this competition, the majors are clearly defending the latter vision. It remains to be seen whether it can prevail without marginalizing legitimate uses of AI, and without allowing commercial rules alone to define what counts as a musical creation.