In late 2025, a strange thing happened to Fave. A version of her song Intentions that she had never recorded, arranged or approved began going viral on TikTok.
An AI collective calling itself Urban Chords had fed her original vocal stems into a model and transformed them into a layered choral arrangement. It was not her voice, exactly, but it was recognisably built from her voice.
She could have sent takedown notices. She could have gone to court. She could have made the case that no one had the right to remake her record without her permission, and she would have been right on the law. Instead, she went into the studio, recorded over the AI arrangement, and released it as an official remix, becoming the first Nigerian artist to formally feature an AI choir on a release. She framed the choice publicly as an act of reclaiming creativity rather than fighting technology and told her fans that Nigerians should be leading innovation, not running from it.
That is one path. It is not the only one. But it is the one the industry should study, because it is the closest thing we have so far to a working template for how an Afrobeats artist can live inside the new reality.
The accusations are the story
In June 2026, Peter Okoye, Mr P, released his comeback single I Can't Look Away. Within days, producers and listeners across X were accusing him of having used generative AI on the record. He has not addressed the claim directly. A month later in July, hip-hop veteran eLDee tha Don previewed I Wish You Well on Instagram and immediately faced the same accusations, with producers pointing to polished backing vocals, unnaturally high notes and the tell-tale rhythmic patterns associated with Suno-generated output. eLDee posted a follow-up video showing the song being mixed, separated vocal stems and all, laughing along with defenders in his comments.
Nobody has proven either case, and it is possible neither is true. But the accusations themselves are the story, because the accusations only exist as a serious cultural conversation when the possibility they describe is realistic. Two years ago, no one would have credibly accused a veteran Nigerian artist of using AI vocals, because the tools were not good enough and the cost of using them was not low enough. Both of those things have changed. The Culture Custodian piece defending the OGs correctly noted that even if AI was used, the ethical question is more about how tools like Suno sourced their training data than about the artist choosing to use them.
The selective-use reality
The way AI is going to be adopted in Afrobeats is not the way the panic suggests. It is not going to be prompt-a-song-and-release-it, at least not by artists who already have careers to protect. It is going to be selective.
If the Mr P accusation is true, the most likely mechanism is that AI was used somewhere in the production chain, and then he recorded his own vocals on top. That is a very specific choice. It cuts out the producer, not the artist. It saves the largest single line item in an independent release budget while preserving the one thing that is legally and commercially most tied to the artist's identity: the voice.
Different artists will make different versions of this trade. Some will use AI for songwriting, letting the model draft hooks they refine by hand and cutting out the outside songwriter fee. Some will use it for beat inspiration, generating twenty reference tracks in an afternoon to figure out what direction they want a project to take. Some will use it for background vocals, harmonies and ad-libs. Some will use it for mastering, because AI mastering delivers competent output in fifteen minutes for less than the cost of one hour with a professional. Some will use it for cover art, lyric videos, TikTok content and playlist targeting. Very few, at the top of the industry, will use it for the lead vocal itself, because that is where the reputational risk lives.
This is what selective use looks like, and it is happening now.
The producer's warning
DJ Neptune, one of Nigeria's most consequential producers, went public earlier this year with a story about being sent a demo he found so impressive he told the person to sign the artist immediately, only to learn afterwards that the whole thing had been generated by AI. He described the experience as having shaken him, and said openly that AI is on the verge of retiring producers and mastering engineers. Media personality N6 has echoed the warning, calling the future of the production and mastering ecosystem scary. Producer Kid-Konnect, contributing to the eLDee debate on X, compared the current wave of AI adoption to the resistance he encountered adopting software-based music production in the early 2000s, when hardware was still king. His point was that the resisters lose that fight every time.
The tell here is that the pushback in Nigeria is not coming from artists first. It is coming from producers, mastering engineers and songwriters, because they are the people whose services AI substitutes for most directly. Artists still have brand, story, likeness, live performance and personal audience relationships that AI does not replicate. Producers and engineers largely have craft, and craft is exactly what generative models are getting better at.
The attention economy argument
Here is the argument on the artist side that deserves to be taken seriously.
According to the IFPI's most recent Engaging with Music global report, the average person spends about 20.7 hours a week listening to music. That is roughly three hours a day. Everyone gets the same 24 hours in a day, and no matter how much music the world produces, the number of hours available for anyone to listen to it does not expand meaningfully.
Now consider the supply side. Deezer reported in June 2026 that it now receives roughly 90,000 fully AI-generated tracks every single day, accounting for more than half of new deliveries to the platform at peak times. That number is not a projection. That is the current daily inflow on one platform.
If the attention available for music is roughly fixed, and the supply of music is exploding, then every stream that goes to an AI track is a stream that did not go to a human artist. Multiply that across months and across the entire streaming economy, and you are describing a fundamental restructuring of how attention and revenue are distributed. That is a legitimate reason for artists to be worried, and it is not a worry that can be waved away by pointing out that AI music is often lower quality. Quality matters less than volume in a discovery environment driven by algorithms and playlist placements.
Why labels do not share the worry
From an artist's standpoint, AI competition is existential. From a label's standpoint, it is neutral, and sometimes beneficial.
A label exists to capture as much of the total music attention pie as possible for the catalogue it controls. If a share of that attention starts flowing to AI-generated tracks, the strategic response is not to fight AI. The strategic response is to make sure the label owns, licenses or has partnership stakes in the AI acts capturing that attention. This is exactly what has already begun. In November 2025, Warner Music Group signed a licensing partnership with Suno, the same generative model most often cited in the Nigerian AI accusations, having earlier sued a similar platform. Warner, Universal and Sony have all signed deals with AI music company Klay to control how their artists' likenesses can and cannot be used in AI outputs.
Read the direction of travel plainly. The majors have moved from suing generative AI companies to licensing them. They have done this because they have concluded that AI music is going to be a permanent feature of the streaming economy, and that owning a piece of it is more profitable than fighting it. Nigerian and other African majors, and the mid-tier labels operating on the continent, are going to reach the same conclusion. Some already have.
The view from inside the industry
The most important Nigerian writing on this topic to appear in 2026 was not published in a music trade outlet or a newspaper. It was a Medium essay posted on 30 July by Muyiwa Awoniyi, Tems' manager and one of the most consequential Nigerian music executives of the current generation. Awoniyi disclosed that an investment opportunity in Suno's Series D funding round, which closed on 3 June 2026, had recently landed on his desk. Whether he took the deal is not the point. The point is that a Nigerian executive at his level is actively weighing an early-stage bet on the generative music platform most often accused of driving the current wave of Afrobeats AI use, and that the framework he sets out for thinking about it maps almost exactly onto the position the American majors have already reached.
His central thesis is that AI's real function in music is not to substitute for artistic vision but to strip away the technical barriers around producing it. He argued that listeners have never really cared about the process behind how their music was made, only about whether it moved them, and that the emotional connection audiences form with music is not automatically closed off to AI-generated work, given how easily people already form attachments to virtual influencers, animated characters and game figures. He predicted a near future of AI-native artists with global audiences, personalised albums generated in real time, interactive music that evolves with a listener's mood, and fans participating as collaborators on records rather than as passive audiences. And he made the point that the biggest winners of the AI shift will not be people who bypass learning music, but the songwriters, producers and artists who understand craft well enough to direct these new tools. "The future isn't human versus AI. It's human with AI," he wrote.
The AI-native precedent
The category of the AI-native artist, an act whose entire creative output is generated rather than performed, is already commercially viable in adjacent genres. Delana Hope, a fully digital gospel artist with an AI-generated voice, reached the Top 10 of the US Digital Gospel Songs chart in 2025 with Something Big. In the Afrobeats-adjacent space, Urban Chords' AI-generated Choir Refix album debuted at No. 43 on the Nigerian Official Top 100 Albums chart in 2025 with roughly 834,000 first-week on-demand streams. The Nigerian producer Eclipse Nkasi released Infinite Echoes, a nine-track fully AI-made Afrobeats album, and documented the workflow openly with international press.
These are not fringe experiments. These are entries on charts that people take seriously. Whether or not any Afrobeats fan personally wants to listen to an AI-generated act, the market data says a share of listeners will, and that share is going to grow.
What the genre should do about It
The right posture for Afrobeats is not defence. Defence loses this fight the same way it lost the fights over Auto-Tune, streaming, TikTok and every other technology cycle that came before. The right posture is leadership.
The labels operating on the continent should be doing three things simultaneously. They should be building internal AI capacity so their artists have access to the tools on terms the label controls rather than terms set by an American platform. They should be writing consent and compensation frameworks that protect their artists' voices, likenesses and stems from being scraped or cloned without payment, ideally before Western regulators write frameworks for other markets that get applied by default to African ones. And they should be developing artist policies that make selective, disclosed AI use a legitimate creative option rather than a shameful secret, because the current culture of hidden use is producing exactly the trust breakdown that the June and July accusation storms revealed.
Artists themselves have a narrower but real choice: to be public about what they are and are not using AI for, so their fans can make informed listening decisions. Fave chose to be public. It worked. The default position of denying and then quietly using is going to age badly.
The debate worth having
AI use in Afrobeats is inevitable because it is already here. It is inevitable because the economics favour it, the tools are cheap, the artists have global-scale content demands that lean production budgets cannot meet, the labels have decided to license rather than fight, and the audience has demonstrated that a chart entry does not require a human singer. None of these conditions are going to reverse.
The debate that is worth having is not whether AI will be used. It is who benefits from its use, whose consent is required, who gets compensated when a voice or a beat or a style becomes training data, and how the industry protects the artists whose work built the sound the models are now generating from.
Fave's move on Intentions was the first sensible Nigerian answer to a real question the industry had been ducking. Awoniyi framed the choice facing the industry as one between spending the next few years defending the way things used to work and spending them building the way things are going to work. That is the right frame. The rest of Afrobeats now has to decide which side of it to be on.
