Grimes

Grimes is a Canadian art-pop and electronic artist who in April 2023 became the first major recording artist to open-license her voice for AI music, offering a 50% royalty split to anyone who released a successful song made with her cloned vocal^[4][5]. The offer became Elf.Tech, a free voice-transformation platform built with CreateSafe that turns any uploaded vocal into her GrimesAI-1 voiceprint^[6]. As of June 2026 the resulting GrimesAI profile on Spotify carries 168 releases from that collaborator commons, while her main profile draws roughly 7.9 million monthly listeners^[2][3].

Background

Grimes is the project of Claire Elise Boucher, born March 17, 1988 in Vancouver and active since 2007, when she began recording in Montreal’s DIY scene around Arbutus Records^[1]. Her 2012 breakthrough Visions (4AD) won the Juno Award for Electronic Album of the Year; Art Angels (2015) debuted at #1 on Billboard’s Top Alternative Albums chart, and Miss Anthropocene (2020) topped Billboard’s Top Dance/Electronic Albums chart^[1]. The standard biography is on Wikipedia.

The Decision. In April 2023, an anonymous AI track faking Drake and The Weeknd — “Heart on My Sleeve” — sent the major labels into takedown mode. Grimes went the opposite direction. “I’ll split 50% royalties on any successful AI generated song that uses my voice,” she posted on April 24, adding: “I have no label and no legal bindings”^[4][5]. That same month she had left Columbia Records^[9]. Within a week the offer was real software.

Career

2010–2020 — Montreal DIY to Miss Anthropocene

Grimes released Geidi Primes and Halfaxa on Arbutus in 2010, then signed to 4AD for Visions (2012), Art Angels (2015), and Miss Anthropocene (2020), building a reputation as a self-producing auteur who writes, performs, and engineers her own records^[1].

2021–2023 — The Columbia interlude

She signed with Columbia Records in March 2021, released the single “Player of Games,” and exited the deal by April 2023, telling the public she had no label^[1][9]. The timing mattered: the open voice offer she made days later was only possible because no label contract restricted her voice rights^[4].

2023 — Elf.Tech and the open voice offer

Six days after the royalty tweet, Grimes announced the Elf.Tech beta: upload or record an a cappella vocal, and the software returns it transformed into GrimesAI-1, an official voiceprint of her voice, as a stem to build a song around^[6]. Rolling Stone covered the experiment under the headline “You Are Grimes Now”^[10]. The first official release followed in May: Kito’s “Cold Touch (feat. GrimesAI)” on Mad Decent, with a source vocal sung by Nina Nesbitt and run through the platform^[11]. In June, TuneCore and CreateSafe formalized distribution: approved GrimesAI collaborations flow to streaming services with a 50% master-royalty split, a GrimesAI artist credit, and Grimes’ sign-off on every release^[7][8].

2023–2024 — The GrimesAI economy

The program scaled into an ecosystem. By fall 2023, more than 200 GrimesAI songs had been licensed for videos, podcasts, and streams through Slip.stream, and CreateSafe added an AI creation-and-distribution partnership with SOUNDS.STUDIO that December^[12]. Releases kept landing on the GrimesAI Spotify profile — Gigamesh’s “The Precipice” (September 2023), 3LAU’s “For You” (January 2024), billed as the first song built from two AI voice clones, and “Hardcore” (May 2024), which became the profile’s most-streamed release^[3]. As of June 2026 the profile holds 168 releases — overwhelmingly third-party Elf.Tech output, not music Grimes made herself — at about 38,000 monthly listeners against her main profile’s 7.9 million^[2][3].

2025–2026 — TIME100, “Artificial Angels,” and Psy Opera

In February 2025 Grimes received a TIME100 AI Impact Award in Dubai, where she called AI “magic” and celebrated its creative potential^[13] — covered on this site in Grimes calls AI ‘magic’ at TIME100 AI Impact Awards. That October she released “Artificial Angels,” a single written from an AI’s point of view, via her Nazgul imprint under exclusive license to Eternal Recurrence^[14]; the AI Musicpreneur breakdown of her exact approach is here. By April 2026 she was teasing a sixth studio album, Psy Opera^[1].

Artistic style and AI workflow

Grimes splices synth-pop, dream pop, electronic, and industrial textures into a sound critics have filed under art pop since Visions; she writes, produces, and engineers her records herself, with science-fiction and AI-consciousness themes running from Miss Anthropocene’s climate-goddess concept to “Artificial Angels”^[1][14].

Her work is classified as Human + AI Hybrid under the SIQA framework — the tier defined by self-voice cloning, where an artist trains a model on their own voice and deploys it as a licensed instrument.

The workflow has two sides. On hers: selective use of the cloned voice inside otherwise human-made records — on “Artificial Angels” she clarified that “the only AI on the song is the voice at the beginning and the end”^[14]. On the public’s: Elf.Tech transforms any submitted vocal into the GrimesAI-1 voiceprint, and approved results are distributed through TuneCore under the 50/50 split^[6][8]. Disclosure is fully on the record — she announced the program herself and labels the AI elements of her own releases publicly^[4][14].

Public image

Grimes is the most visible AI advocate among established recording artists — a stance institutionalized by the TIME100 AI Impact Award in 2025^[13]. The position is more conditional than the headlines suggest: she has said she is opposed to AI music “in some forms” and has criticized current AI music apps as “slop oriented,” while defending consent-based, artist-controlled deployments like her own^[14]. Her open-voice offer remains the reference event for consent-based voice cloning — the counter-model to the unlicensed deepfakes that triggered it^[5][10].