Grimes’ new “Accelerando” video arrives at a moment when audiences increasingly treat artificial intelligence as an authorship question, not merely a production tool. The most useful part of the release may be its credits: the video identifies more than 70 people across performance, choreography, makeup, production and visual effects, while also naming “AI artists.”
According to a representative for Grimes, artificial intelligence was not used to record the song or its vocals. AI-generated elements were selectively incorporated into a wider visual-effects process that also used live performers, physical materials, blue-screen filming and conventional computer-generated imagery. The production reportedly took two to three months, and the representative described AI as a supplementary element within an artist-led pipeline.
The credit list is the real creative document
The disclosure matters because “made with AI” is too blunt a label for a hybrid production. It does not tell a viewer whether a model generated an initial texture, altered footage, supplied finished scenes, imitated a recognizable style or replaced work that would otherwise have gone to a human specialist. Nor does “human-made” explain how much software shaped the final image.
“Accelerando” offers a more practical model: separate the music from the visuals, distinguish live action from generated elements, and preserve the names of the people whose judgment made the finished work possible. That does not settle questions about training data, consent or copyright. It does, however, make the process more legible and gives audiences a basis for asking better questions.
What artists can take from this
Artists working with generative systems can begin by keeping a simple production record. Note which tools were used, where generated material entered the workflow, what source material was supplied, who selected or transformed the output, and which collaborators carried responsibility for the finished work. Contracts should also define whether a contributor is providing direction, model operation, compositing, performance, design or another form of labor.
Credits should not collapse every contribution into the word “AI.” A useful credit architecture names the human roles first and identifies the system only where it affected the work. When a model’s training or licensing terms are material, those should be recorded separately from the public-facing credit list.
Grimes has made artificial intelligence part of the conceptual frame for her forthcoming album Psy Opera, and she has previously invited experimentation with a synthetic version of her voice under a royalty-sharing proposal. That history makes this release unusually visible. Yet the broader lesson applies far beyond music videos: transparency becomes meaningful when it describes the actual chain of creation rather than offering a single reassuring label.
ATIM on Artificial Intelligence
ATIM approaches artificial intelligence with curiosity, independence, and balance. We explore its extraordinary potential to empower artists while examining the questions it raises around authorship, copyright, consent, creative identity, transparency, and human expression.
We are neither advocates for AI nor opponents of it. Our focus is how the technology is used—and what that means for artists.
Above all, our commitment is to artists. We look at every development through the lens of what can support, protect, and empower the creative community—identifying opportunities, examining potential risks, and providing artists with the information they need to make informed decisions about their work, their rights, and their future.