There is a moment during every World Cup that transcends the sport itself — when the stadium, regardless of where it sits in the world, becomes something closer to a collective hallucination. Strangers, nationalities scrambled by migration and accident, find themselves chanting the same syllables, swaying to the same rhythm. Music has always been the mechanism that makes this possible. From “Waka Waka” in Johannesburg to “La’eeb” reverberating across Doha, the tournament’s anthems carry a weight that FIFA contracts cannot fully account for. They become, in some ineffable way, everyone’s.
This summer, for the first time, the crowd is writing back.
Months before the opening whistle in Mexico City on June 11, fans across dozens of countries had already produced hundreds of World Cup songs — not through studios, not through label deals, but through AI music generation tools that required nothing more than a text prompt and a few seconds of processing time. By the time Deezer, the only major streaming platform currently flagging AI-generated content, published its findings in early June, the numbers were striking: more than 270 tracks titled “World Cup 2026” had been uploaded to the service, with over 70 percent detected as machine-made. Among French-language releases using the title “Coupe du Monde 2026,” the figure climbed to 86 percent. In Brazil, 71 percent. The pattern, Deezer noted, held across markets with minimal variation. It was not a regional curiosity. It was something systemic.
The track that arguably started the wave was “Imbattables,” a French fan anthem released in February by an artist going by Crystalo — billed on Spotify, with a certain pride, as France’s “premier AI musical creator.” The formula was simple: call-and-response, player names chanted in sequence, a propulsive beat that loops cleanly across TikTok’s fifteen-second format. It accumulated millions of views. Fans who didn’t speak French shared it anyway. The song had the texture of something that had always existed at the edge of match-day playlists, which was partly the point. AI, it turns out, is extremely good at making things that sound familiar.
That familiarity cuts both ways. Part of what these tools do — platforms like Suno, Udio, and others that have made AI music creation a genuinely mass phenomenon — is compress the entire history of recorded sound into something retrievable by prompt. A user who asks for “a stadium anthem with a Brazilian samba feel and a chorus that can be chanted” is not composing so much as curating from an enormous, largely unlicensed library of musical DNA. Jason Palamara, a music technology professor at Indiana University, has been direct about the problem this creates: there is, he says, a fundamental lack of clarity about how artists are credited when their copyrighted work is used to train these models. The song that emerges from the machine carries within it traces of songs that were never licensed to be there.
The music industry is fully aware of this, and the legal and lobbying infrastructure now arrayed against AI companies reflects how seriously the threat is being taken. CISAC, the global body representing over five million music creators, commissioned a landmark study whose conclusions were sobering: by 2028, AI substitution could put roughly 24 percent of music creators’ revenues at risk — approximately ten billion euros. Streaming, historically the sector’s growth engine, faces a particularly acute exposure; AI-generated music is projected to account for around 20 percent of traditional streaming platform content by the end of the decade. The UK government’s decision in March to abandon proposals that would have allowed AI companies to train on copyrighted work without permission — proposals that attracted over ten thousand consultation submissions, with only three percent in favour — offered the industry a rare legislative win. But the underlying economics remain unchanged. Deezer is already ingesting roughly 75,000 fully AI-generated tracks every single day, representing 44 percent of its daily uploads. The flood is not coming. It is here.
Against this backdrop, FIFA’s choice of official anthem looks, in retrospect, almost designed to accelerate the problem. “Lighter,” the first single from the tournament’s official album, pairs American country singer Jelly Roll with Mexican regional star Carín León. Produced by Cirkut — fresh from being named Producer of the Year at the 2026 Grammys — the song is technically accomplished, the collaboration culturally deliberate. FIFA framed it as a bridge between the tournament’s three host nations, a sonic handshake between Nashville and Monterrey.
The internet was not moved. The backlash was swift and, in its specificity, instructive. Fans called it a Dodge Ram commercial. A music critic’s description — “a Bible Belt ballad without a pulse” — circulated widely. The hashtag #NotMyWorldCupSong trended before the chorus had finished playing on the official livestream. What the song lacked, according to its critics, was not craft. It was the quality that made “Waka Waka” an event: the sense of a world genuinely gathered, of music that could belong to someone arriving from Dakar or Seoul or Warsaw with no particular investment in either country music or regional Mexican. The cup is global; the anthem felt local.
The gap this created was the gap that AI rushed to fill. When the official song fails to inspire, fans have historically made their own — terrace chants, parody versions, bedroom recordings circulated on cassette and later on YouTube. What has changed is scale and speed. The same week “Lighter” dropped, creators were already prompting alternatives. Some of those alternatives accumulated more genuine enthusiasm than the official release. Several fans expressed, without apparent irony, that they found the machine-made tracks more emotionally satisfying than the FIFA-sanctioned one.
This preference — aesthetic, not ideological — is where the conversation gets genuinely uncomfortable. One AI music creator working on a Brazilian fan track told AFP that he views AI as an assistant, a tool in a human creative process rather than a replacement for it. “What I see happening now is more about people following a trend or trying to recreate a feeling,” he said, adding that artistic emulation has always existed in music. He is not wrong. Pastiche is as old as music itself. The question is whether pastiche built on an unlicensed foundation is something the industry can sustain, or whether the “trend” he describes is quietly redistributing the economic value of human creativity toward platforms and prompts.
The platforms are navigating this unevenly. Deezer remains, as of now, the only major streaming service that detects and labels AI-generated music at scale, automatically excluding such tracks from algorithmic and editorial recommendations — a measure its chief innovation officer says ensures that most AI World Cup songs will never accumulate meaningful streams. Apple Music has announced a Transparency Tags system, allowing distributors to flag AI-assisted content through release metadata. The broader direction of travel is toward disclosure rather than prohibition. What remains unresolved, and fiercely contested, is the upstream question: what happens to the artists whose work trained the models in the first place.
There is something historically appropriate about this argument erupting around a World Cup. Football has always been the sport where the relationship between institutional authority and popular culture is most visibly strained. FIFA controls the tournament with a thoroughness that no other governing body attempts; it commissions the songs, owns the footage, enforces the logo. And yet the tournament belongs, viscerally and undeniably, to the people who show up — who create the noise, the colour, the moments that television captures but does not produce. The tension between what FIFA owns and what the crowd feels has always been the subtext of the competition.
AI fan songs are, in one sense, that tension finding a new outlet. The crowd has always wanted to write the anthem. Now, with sufficient bandwidth and a free account on a generative platform, it can. Whether the industry finds a way to compensate the musicians whose life’s work made that possible — whether the “thorny Rubicon,” as one expert put it, actually gets crossed — will determine whether what has emerged this summer is a genuine democratization of music or simply a more sophisticated form of taking.
For now, in stadiums across the United States, Mexico, and Canada, both kinds of songs are playing. Some were made in recording studios with Grammy-winning producers. Some were made in the time it takes to write a sentence. The crowd, as it always has, will decide which ones it remembers.
Enjoyed This Story? Subscribe to EntertainLens for more entertainment insights.
