| An open letter |
| Verified, Not Flagged. |
| September 14, 2026 |
| |
Last week, we launched the SIQA Verified Registry. Over the past few months, we've received many questions about our stance on the industry developments in AI music, and where SIQA fits into them. This letter is our answer, and an introduction to the infrastructure behind the next phase in the future of transparency. |
|
|
| Something has shifted. |
| Between January and August of 2026, the AI music conversation moved: visibly, publicly, and in some cases, courageously. Artists who had quietly been creating with AI tools began saying so out loud. Platforms began acknowledging that AI music existed in their catalogs and that they had a responsibility to do something about it. Industry organizations began convening, proposing frameworks, and attempting to define what AI music actually is. That is movement. And it deserves to be acknowledged. |
| SIQA is encouraged by the direction of this movement. The fact that transparency in AI music is now an active institutional conversation, not just a fringe debate, is exactly what this category has needed. The artists who have publicly advocated for openness about their creative process, at real professional risk, have helped move that conversation faster than any white paper or industry summit could. |
| But encouragement is not endorsement of every solution being proposed. And the framing of many of those solutions reveals a foundational assumption worth examining. |
| Our stance on labeling: the trap of the single AI-Generated tag |
| Much of the industry's current approach to AI music transparency is built around a single question: is this music AI-generated or not? It is an understandable question. It is also, in most cases, the wrong one. |
| SIQA's mission is not to determine what is AI and what is real. It is to provide transparency about how AI was used, and to give proper credit and context to the human contribution behind every piece of music in this ecosystem, whether that contribution is a producer directing an AI tool through a series of deliberate prompts, a songwriter whose lyrics anchor an AI-generated production, or a recording artist who used an AI tool to reproduce and extend their own voice. |
| The difference between those three artists is not the presence or absence of AI. It is the nature, degree, and intent of human creative involvement. A framework that collapses all three into a single "AI-Generated" label does not create transparency. It creates a ceiling where the floor should be. |
| A label that flags music as "AI-Generated" without distinguishing between a fully machine-produced track and a record anchored in human songwriting, production direction, or vocal performance tells a listener almost nothing useful about how the music was actually made. And for the artist on the receiving end of that label, it can reduce a specific, deliberate creative process to a designation that carries, fairly or not, a significant cultural stigma, regardless of how much of themselves they put into the work. |
| "Is this AI?" and "how was AI used, and what did the human bring to it?" are two very different questions. And the answer to the second one is where the artist lives. |
| On industry definitions: the right conversation, the incomplete conclusion |
| Several traditional and legacy organizations across the music industry, including the IFPI and The Recording Academy, have made meaningful efforts to engage with the question of what AI-assisted music means, and what obligations, if any, it creates for the artists, platforms, and organizations that touch it. SIQA welcomes those conversations. The willingness to try to define something difficult, in public, with real institutional weight behind it, is an act of good faith that the industry needs more of. |
| The current proposed frameworks, however, are incomplete in ways that matter. |
| A definition of "AI-assisted" that draws a bright line at the point of AI involvement, without accounting for the degree, nature, or creative intent of that involvement, risks doing the very thing it intends to prevent: making human artistry invisible. An artist who writes every lyric, shapes every production decision, performs on the track, and uses an AI tool at one stage of the process is a fundamentally different creative actor than one who inputs a prompt and selects from generated options. Treating them identically under the same definition does not protect artists. It erases the distinction between them. |
| Awards bodies, credentialing organizations, and recognition institutions play a consequential role in defining what counts as human enough to be recognized. Those determinations shape what gets awarded, what gets promoted, and ultimately what gets made. That is precisely why a definition that excludes more than it includes is not a conservative position. It is a consequential one. |
| On SIQA's framework: designed for the artist, not the category |
| When we built the SIQA Classification Framework (AI-Assisted, Human + AI Hybrid, and Fully AI-Generated), we did so with a specific intention: to reflect the actual creative reality of artists working in this space, and to ensure that human contribution is credited and conveyed at every level of that reality. |
| AI-Assisted recognizes that a human fulfilling one or more core creative roles alongside AI tools is a meaningfully different creative actor than one who has stepped back from the process entirely. Human + AI Hybrid acknowledges the specific and growing practice of self-voice cloning, an artist reproducing their own voice through AI, as a distinct creative act that carries the artist's own identity, likeness, and creative intention directly into the output. Fully AI-Generated creates space for music that is openly machine-produced and transparently disclosed, with the human role centered on direction, curation, and selection. |
| These distinctions exist not to create a hierarchy, but to give every artist in this ecosystem the most accurate, most dignified description of what they actually did. No framework is perfect. SIQA's is not the final word. But the process that produced it, first-party creator disclosure, formal verification, and the genuine effort to let artists define their own work rather than have it defined for them, is the foundation any useful definition of AI music will need to build on. |
| A seat at the table |
| What this moment needs, and what is notably absent from most of the conversations currently happening, is a neutral institution with verified data, a track record of working directly with AI music creators, and no financial stake in a particular outcome. SIQA is that institution. We are not a platform with licensing incentives. We are not a rights organization with legacy interests to protect. We are not a trade body whose membership has a collective position to defend. We are a data institution built around one purpose: understanding AI music from the inside out, with verified, first-party evidence, and without a predetermined conclusion. |
| It is worth noting that some of the most significant institutions now working on AI music classification have said publicly that no single organization can address this challenge alone, and that a collaborative approach is essential to accurately understanding, tracking, and measuring AI-generated music. SIQA agrees entirely. First-party, creator-disclosed classification data is something no detection-based system can replicate. It is precisely what the industry's emerging frameworks are missing. And it is what we have been building since January 2026. |
| We are actively seeking to connect and collaborate with the organizations working to solve this. We are ready to bring what we have built to the table. And we believe that the most accurate, most artist-respecting picture of AI music will only emerge when multiple perspectives and methodologies are working together rather than in parallel. |
| Introducing: The SIQA Verified Registry |
| We are happy to announce the launch of the SIQA Verified Registry. It extends the work we began with our charts: from ranking what's rising to verifying how it was made. |
 |
| The SIQA Verified Registry lets artists register their work and establish provenance up front, turning verification into a credential artists carry rather than an accusation they have to defend against. It goes beyond the generic "AI-Generated" label to classify each submission under the SIQA Classification Framework as Fully AI-Generated, AI-Assisted, or a Human + AI Hybrid. For artists, it is proof of process. For fans, it is context for the music they love. For the industry, labels, distributors, DSPs, and rights organizations, it is a trusted way to check a track's provenance. The SIQA Verified Registry is live now at registry.thesiqa.com. |
 |
| Several platforms have already begun building disclosure mechanisms of their own, from distributor-level AI disclosure fields to DSP credits systems. We see those efforts as complementary, not competitive. The more places artists are asked to disclose how they made their music, the more important it becomes to have a single, verified, portable record of that disclosure. That is what the SIQA Verified Registry provides. |
 |
| As DSPs, distributors, and music tech platforms seek to identify AI use in music beyond the generic "AI-Generated" tag, SIQA is open to data partnerships with those who inquire, making its registry data available to power those efforts. SIQA's registry is also accessible via MCP, allowing AI agents and partner systems to verify tracks and query provenance data directly. |
| Rather than policing or labeling tracks as suspect after the fact, we built a system that equips artists to verify at the source. The prevailing industry response to AI music has been reactive: identify tracks after release, then flag, dispute, or remove them. Our intention is different. |
| To Verify, Not Flag. |
| Flagging starts with suspicion. Verification starts with the artist. We built the registry to start there, and we built it for anyone who wants to start there with us. |
 |
| President, The Sonic Intelligence Academy (SIQA) |
| & The SIQA Team |
| Building for AI transparency? Let's talk. |
| SIQA is open to data partnerships with DSPs, distributors, rights organizations, and music tech platforms looking to go beyond the generic "AI-Generated" tag. Registry data is available via API and MCP. |
|
|
| Read the SIQA Mid-Year 2026 AI Music Intelligence Report |
| Everything we learned across the first half of 2026: the data behind the charts, the classification breakdown, and where the category is heading. |
|
|