Global music companies propose chart rules for AI-developed recordings

Research report date: 30 July 2026

DAOrecords Signal publication: 30 July 2026 at 03:58:52 UTC

Recovery-window disclosure: The canonical joint announcement is dated July 29, 2026 but exposes no exact publication time. Conservatively normalising it to 2026-07-29T00:00:00Z places it 27 hours, 58 minutes and 52 seconds before this research run. It is therefore included through the 24–72-hour recovery lane and should not be represented as breaking coverage from the latest 24-hour window.

Status: Confirmed industry proposal; not an adopted chart rule

Linked record: DAOR-SIGNAL-20260730-001

Factual reporting

Believe, BMG, Concord, Dirty Hit, Glassnote Records, HYBE, Mom+Pop Music, Partisan Records, Sony Music, Universal Music Group and Warner Music Group have jointly proposed global principles for deciding whether recordings developed with generative-AI services should qualify for official music charts.

The proposal states that an AI-developed recording should not be chart-eligible where there is reason to believe that it fails any of six conditions:

1. Every generative-AI service used to develop the recording is properly authorised and lawful.
2. The recording is substantially human-made.
3. The recording does not raise stream- or chart-manipulation concerns.
4. The recording complies with applicable copyright, related-rights and personality-rights law.
5. Distribution of the recording does not breach the terms of the AI service used.
6. The use of generative AI is appropriately signalled to consumers on downstream services in accordance with applicable law or industry labelling standards.

The companies describe the principles as a roadmap for chart compilers, industry bodies and affiliated stakeholders. They have offered to discuss the framework and support implementation.

The announcement does not identify a chart compiler that has adopted the proposal, establish an implementation date, define “substantially human-made,” or specify what evidence would prove that a model or service is authorised and lawful.

The proposal follows the music community's July 10 voluntary labelling framework distinguishing “AI-Generated” from “AI-Assisted” recordings. The new announcement states that recordings meeting the “AI-Generated” definition would not be considered chart-eligible under the proposed principles.

Why it matters to the music industry

The proposal attempts to connect AI provenance, creator rights, human contribution, disclosure and streaming integrity to a concrete market outcome: eligibility for official chart recognition.

This is significant because chart eligibility can affect visibility, promotion, reputation and commercial momentum. The framework would make AI governance relevant not only at creation and distribution, but also at the point where industry institutions certify popularity and cultural impact.

The coalition is unusually broad. It includes the three major recorded-music groups alongside major independents and international companies. That does not make the proposal binding, but it gives the framework substantial industry weight.

The proposal also rejects a simple binary test of whether AI was used. AI-assisted work could remain eligible where the service is lawful, the work is substantially human-made, rights are respected, disclosure is supplied and manipulation concerns are cleared.

The unresolved issue is evidence. Chart compilers would need reliable data from AI services, distributors, labels and DSPs to assess authorisation, human contribution, rights compliance, disclosure and fraud risk. An audio detector cannot establish all of those conditions.

Supporting research: detection is not provenance

The July 28 preprint Finding the noise: Zero-shot AI Music Detection proposes methods for detecting synthetic music from previously unseen generators by combining artifact extraction, non-negative matrix factorisation, one-class classification and clustering.

The paper is relevant because chart and catalogue systems will encounter outputs from new or changing models. However, its reported detector performance does not provide a complete chart-eligibility test.

Detection can indicate that a recording may contain synthetic artifacts. It cannot by itself prove whether the service was licensed, whether the recording is substantially human-made, whether voice or personality rights were authorised, whether disclosure metadata is accurate, or whether streams were manipulated.

The research is also a preprint rather than an adopted technical standard. Its datasets, generator coverage and treatment of hybrid recordings limit its direct use as a governance control.

DAOrecords analysis

DAOrecords should represent chart-readiness as an evidence bundle rather than a single AI-origin flag.

For a release developed with generative AI, the ReleaseOps and Rights + Metadata layers should be capable of preserving:

• the AI service, model and version used;
• evidence that the service and relevant model use are authorised and lawful;
• the human contributors and the nature of their creative contribution;
• composition, master, neighbouring-rights, voice and personality-rights authority;
• the AI-origin classification, including whether the work is AI-assisted or AI-generated;
• downstream disclosure fields supplied to distributors and DSPs;
• the AI service's applicable terms and evidence that distribution is permitted;
• stream-integrity and chart-manipulation screening;
• provenance evidence, dates, supporting documents and responsible reviewers;
• the chart body, applicable rules, decision status and any appeal or correction history.

A DAOrecords Child Vessel should not assert that a release is “chart eligible” merely because a distributor accepted delivery or a detector classified the audio as human-made. Chart eligibility is an external governance decision and may vary by chart, jurisdiction and rule version.

The proposal should be added to the Parent policy watch. DAOrecords should monitor whether chart compilers adopt, modify or reject the framework and whether distributors introduce new metadata, warranties or evidence requirements in response.

The future governance challenge is versioning. A recording may be eligible under one chart body's rules and ineligible under another's. DAOrecords should preserve the exact rule set, jurisdiction and decision date instead of storing chart eligibility as a timeless universal property.

Assessment

• Impact level: High
• Confidence: High that the named companies jointly issued the proposal and stated the six eligibility conditions
• Source status: Proposal; no global chart rule has been adopted
• Affected components: AI Rights Profile; Rights + Metadata; ReleaseOps; Streaming Integrity; Evidence Preservation; Ecosystem Relations
• Canonical source: Sony Music joint announcement (opens in a new tab)
• Canonical source date: 29 July 2026
• Supporting research: Finding the noise: Zero-shot AI Music Detection (opens in a new tab)
• Commercial data value: High for AI-origin disclosure, chart governance, rights verification, release eligibility and streaming-integrity controls
• Primary limitations: No participating chart compiler, implementation date, evidence standard, definition of substantial human contribution, enforcement process, appeal mechanism or cross-jurisdiction adoption plan has been announced.

Daily synthesis

The proposal moves AI-music governance closer to an institutional decision point. Rights compliance, human contribution, disclosure and fraud screening would influence whether a recording can receive official chart recognition.

For DAOrecords, the core requirement is an auditable release evidence bundle. A detector score or generic “AI-assisted” tag is insufficient. Chart-readiness requires verifiable service authorisation, contribution records, rights authority, disclosure metadata, terms compliance and integrity screening tied to a specific chart rule and decision date.

Record index