Published September 30, 2024 · By CampaignsLive · Industry
Through 2024, generative music made the transition that generative video would not complete until later. The category went from research demonstration and novelty output to production-grade work that real audio teams could integrate into real projects. Two names — Suno and Udio — became the dominant references in the conversation, displacing the slower-progressing Google MusicLM and Meta MusicGen efforts that had set the pace in 2023.
The implications for brand audio production took the rest of 2024 to become visible. By the end of the year, the pattern of where generative music fit and where it did not had started to resolve.
What changed in 2024
Three things, in roughly the order they became visible.
The output became musically coherent at full song length. Through 2023, the generative music tools produced clips — fifteen to thirty seconds — that demonstrated the model could generate plausible musical material in a given style. Through 2024, the tools moved to full-song generation: two-to-four-minute outputs with verse-chorus structure, dynamic variation, and the kind of arrangement that audiences expect from commercial music. The shift was the most visible technical advance of the year.
Vocal generation became convincing. Through 2023, generative tools could produce instrumental music at acceptable quality but struggled with vocal performance. Through 2024, Suno and Udio in particular produced vocal performances that crossed the casual-listener plausibility threshold. The vocals were not perfect — careful listeners could identify markers — but they were good enough that the output read as music rather than as demonstration.
Genre and style range expanded. The 2023 generative music tools had been most reliable on a narrow set of styles — typically electronic, ambient, or production-music-adjacent genres. Through 2024, the tools became credible across pop, rock, country, hip-hop, classical-adjacent, and the long tail of specific subgenres. Brand audio production, which spans almost every genre depending on the campaign, gained access to a meaningfully broader range of generative material.
What this changed for brand audio production
Three use cases shifted from experimental to operational through 2024.
Background music for brand films and social content. The category of generic-feeling but appropriately-toned background music that brand films routinely need — the “thing that sounds like it belongs here” rather than the named-artist license — moved decisively into the generative pipeline. The economic case was strong: the cost of generating appropriate background music was a fraction of the cost of licensing comparable material from stock-music libraries.
Concept and animatic audio. The audio for pre-production storyboards, animatics, and stakeholder reviews moved into the generative tools. The work that had historically been a placeholder — generic library music selected to give a rough sense of the eventual scored audio — became generative output specific to the project. Stakeholders could react to audio that approximated the final intent earlier in the production cycle.
Lower-tier branded content. The branded content that ran at lower production budgets — social-channel content, internal communications, sales material, regional brand content — increasingly used generative music in place of licensed material. The shift compressed the cost of branded audio production substantially in this category.
Where generative music did not move
Two categories remained traditional production.
Hero campaign music. The music for the year’s major brand campaigns — the work that audiences would associate with the brand for the next eighteen months — continued to come from commissioned composers, licensed catalogue, or named-artist work. The reasons were partly aesthetic (the hero music has to be specifically right rather than appropriately generic) and partly relational (the brand-artist relationship has value beyond the per-project output).
Named-artist endorsements and partnerships. The brand-artist endorsement deals that anchor major campaigns — the named musician, the recognizable voice, the specific cultural association — remained the work of actual artists. Generative tooling did not, and likely will not, replace this category. The category exists because the artist’s recognition is the value; synthesizing the recognition undermines the value.
The rights and labor questions that came along
The same questions that had shaped the image-side and visual-likeness conversations through 2023 reached generative music through 2024.
Training-corpus concerns intensified. The major music labels — Universal, Sony, Warner — pursued more aggressive enforcement of training-corpus rights through the second half of 2024, with the Suno and Udio lawsuits as the most visible cases. The legal exposure on generative music tools trained on commercially-released music corpora has not been definitively settled; the litigation will likely take years to resolve.
Voice synthesis policies converged. The unauthorized synthesis of identifiable artists’ voices — the dynamic that the Drake/Weeknd episode had crystallized in 2023 — became more aggressively policed by the platforms. By the end of 2024, the leading generative music tools had built in restrictions on synthesizing specific named artists, even when users explicitly requested them. The restrictions are imperfect but materially more aggressive than they had been a year earlier.
SAG-AFTRA-style labor language extended to musicians. The American Federation of Musicians and adjacent organizations began advocating for music-industry equivalents of the AI clauses that the SAG-AFTRA negotiations had established for performers. Progress through 2024 was slower than on the actor side but moving in the same direction.
What this leaves for brand audio teams
Three working positions emerged through 2024.
Generative music is appropriate for production-tier work below the hero level. Background music, animatic audio, social and internal content, lower-budget branded content. The category boundary is reasonable to operate inside; brands that have done so have absorbed real production-cost savings without obvious brand-equity cost.
Hero campaign music remains traditional. The named work that defines the campaign — composer commissions, licensed catalogue, artist partnerships — continues to be the right tooling for the equity-driving work. Generative tooling is not the right answer for this category, and brands that have tried have generally absorbed reputational cost from the choice.
Rights documentation is required for the work that is generative. The training-corpus and voice-synthesis litigation are not resolved. Brands using generative music in commercial production should maintain provenance documentation for the work, including which tool was used, what inputs were provided, and what rights documentation the tool’s terms grant. The infrastructure for this is more mature on the image side than on the music side; brands running both should not assume the music-side infrastructure is at the same level.
For the related visual-likeness conversation that paralleled this through 2023 and 2024, see Drake, the Weeknd, and the 2023 AI Music Controversy and The SAG-AFTRA AI Clauses.