For years, musicians have argued that companies should not be able to train commercial generative systems on music without asking the people who created and own it.

LANDR is proposing a different model. Eligible artists can opt in, allow qualifying recordings to enter AI-training datasets, receive an advance when selected for a promotion, and share in revenue generated from the licensing and exploitation of those datasets and related AI products.

That is a meaningful development. It recognizes that music used to build AI systems has economic value and that rightsholders should participate in that value.

It also creates a more difficult question:

This article examines the actual business decision. It is not an endorsement of LANDR, and it is not an argument that every artist should reject the program. It is a plain-language review of what creators receive, what they authorize, which works qualify, what opting out can accomplish, and what the public terms still leave unclear.

What LANDR is offering

LANDR’s Fair Trade AI program allows qualifying LANDR Distribution users to opt eligible recordings into datasets that LANDR or outside AI collaborators can use to train models and develop AI products.

The current promotion advertises an initial $1 million advance fund and a $5 recoupable advance for each approved eligible track accepted under the offer. LANDR’s current promotional material also advertises an artist pool equal to 25% of net licensing revenue.

LANDR says participating artists retain ownership of their original music. Participation is voluntary, and artists who have not accepted an advance can generally opt out of future licensing. Artists who accept an advance face a required participation period tied to recoupment or a maximum of two years.

The $5 payment is an advance—not a bonus

The most important word in the offer is recoupable.

An advance is money paid before the artist’s share of program revenue has accumulated. Future amounts allocated to the artist under the Fair Trade AI program are then used to recoup that advance. Regular Fair Trade AI payouts begin only after the accepted advance has been fully recouped.

Simple example

LANDR’s terms say the company will not normally pursue the artist personally for an unrecouped balance. The advance is recouped from money payable under the Fair Trade AI section, not from ordinary distribution royalties. However, an artist who wants to leave early may need to voluntarily repay the remaining balance.

Not every track in a LANDR account qualifies

The offer applies only to approved Qualifying Recordings. Uploading or distributing a song through LANDR does not automatically make it eligible for the program or guarantee a $5 advance.

LANDR reserves the right to verify, curate, accept or remove recordings. It also decides which datasets will contain each accepted recording. A track can therefore pass several different gates:

1. It exists in the artist’s LANDR distribution catalog.
2. The artist opts it into Fair Trade AI.
3. It meets LANDR’s full-rights requirements.
4. LANDR approves its quality and eligibility.
5. LANDR selects it for one or more datasets.
6. An AI partner licenses or uses a dataset containing it.

Creators should not multiply their total catalog size by five dollars and assume that amount is guaranteed.

Who qualifies—and why the standard is stricter than “I made the song”

LANDR’s terms require the participant to own and control all relevant rights in the recording, including:

• 100% of the sound recording or master rights;
• 100% of neighbouring and performer-related rights;
• 100% of the publishing rights;
• mechanical, performance and lyric rights;
• worldwide authority to grant the required licence;
• the ability to confirm that LANDR, its partners and users will not owe another party for the permitted use.

LANDR expressly excludes categories including cover songs, sampled works, some works containing public-domain material, and recordings represented in whole or in part by a publisher, record label or collecting society.

Creators should confirm all of the following

• Who owns the master?
• Who wrote the melody and lyrics?
• Was another producer or musician promised a share?
• Did a featured singer or instrumentalist grant the necessary performer rights?
• Does the song contain samples, licensed loops or third-party recordings?
• Is the composition administered or represented by another company or society?
• Can every required warranty be made truthfully and in writing?

Important clarification for Suno creators

This program should not be treated as an opportunity to submit a catalog of Suno-generated songs for another AI company to train on.

LANDR requires complete control of the master, publishing, neighbouring and related rights necessary to grant its broad AI-development licence. A Suno paid-plan commercial-use permission helps a creator release and monetize eligible outputs under Suno’s contract, but commercial-use permission is not the same as independently controlling every underlying right and being able to make LANDR’s full ownership warranties.

That distinction is already central to the Jack Righteous guide to owning and commercially using AI music.

The LANDR offer is primarily relevant to musicians who can demonstrate complete control of qualifying recordings and compositions—not creators trying to relicense finished Suno generations as training assets.

The advertised 25% and the current terms do not match

This is the most important unresolved issue in LANDR’s public presentation.

LANDR’s promotional Fair Trade AI page currently advertises a 25% share of net licensing revenue. LANDR’s currently published Terms of Service still describe a 20% pro-rata share.

The documents also appear to describe allocation differently. Promotional explanations have emphasized participation by recordings or tracks, while the current terms say the pro-rata share is based on the total seconds of audio contributed.

This discrepancy may simply reflect an updated promotion that has not yet been consistently incorporated across all public documents. It should not automatically be characterized as deception. It should, however, stop artists from relying on the 25% headline without reading the agreement presented at opt-in.

Twenty-five percent—or twenty percent—of what?

Even after the percentage is resolved, the artist’s share is calculated from net revenue, not gross revenue.

LANDR’s terms allow deductions such as taxes, refunds, discounts, rebates, payment-processing fees, credits and chargebacks before the artist pool is calculated. The exact value of the share therefore depends on information that is not contained in the headline:

• the gross value of each dataset or AI-product agreement;
• which deductions apply;
• the size and duration of the licensed dataset;
• the amount of qualifying audio contributed by each artist;
• whether the agreement is one-time or recurring;
• whether downstream AI-product revenue is included;
• how revenue is attributed when a product uses multiple datasets or models.

An illustrative calculation—not an earnings forecast

Suppose a partner agreement produced $100,000 in qualifying net revenue and the artist pool were 25%. That would create a $25,000 artist pool. If an individual creator supplied one ten-thousandth of the qualifying audio under the governing allocation formula, that creator’s allocation would be approximately $2.50 before considering any outstanding advance.

That example does not represent an announced LANDR deal. It simply shows why the size of the dataset and the allocation method matter as much as the headline percentage.

What rights does an artist grant?

LANDR’s terms describe a broad, non-exclusive, transferable, sublicensable, worldwide licence. Subject to the program’s opt-out limitations, the licence is described as irrevocable for uses already completed or rights already granted.

The licence allows LANDR and its AI collaborators to use qualifying recordings to:

• reproduce, store and convert the audio;
• modify, alter, remix and create derivative material;
• train, improve, test, validate and evaluate neural networks;
• develop AI models and commercial AI products;
• create or enable users to create commercially exploitable outputs;
• conduct AI research and development;
• use the material for internal business purposes connected to AI systems and products.

The artist keeps ownership of the original qualifying recording. That is important—but ownership retention does not make the licence narrow.

Who owns the models and outputs?

LANDR’s terms say that, as between the artist and LANDR, ownership of the AI models, related data and generated outputs remains with LANDR or the applicable AI collaborator, subject to future changes in copyright law.

The outputs described in the terms can include full-length tracks, samples, one-shots, hits, beats, loops, stems, compositions, lyrics and similar material. LANDR and its collaborators may exploit or allow users to exploit those outputs without additional payment beyond the compensation specified in the Fair Trade AI arrangement.

That leads to the central value question:

There is no universal correct answer. Some artists will see a licensed dataset as a reasonable new use for dormant catalog material. Others will believe the advance and uncertain revenue share do not adequately reflect the downstream value.

Does the AI model contain a playable copy of the song?

LANDR explains that AI systems learn musical patterns and relationships rather than functioning as a searchable locker containing a playable copy of every training recording.

That explanation should not be expanded into guarantees LANDR has not made. It does not automatically prove that:

• no model can memorize source material;
• no output can resemble an input;
• no copyright dispute can arise;
• no artist’s style or identifiable musical traits can be imitated;
• every AI partner uses the same technical safeguards.

The practical decision is simpler: artists must decide whether they are comfortable allowing their music to contribute to systems designed to generate new commercial audio from learned musical relationships.

Time-limited and perpetual licences are not the same

LANDR’s public Fair Trade AI explanation distinguishes between partner arrangements that may be time-limited and arrangements that may allow a partner to retain a delivered dataset indefinitely.

The public materials do not clearly give artists the ability to approve each partner, reject individual datasets, accept time-limited licences while refusing perpetual ones, or set a minimum price for a particular use.

Unless LANDR provides additional controls during enrollment, opting into the program means trusting LANDR to select the datasets, partners and deal structures.

What opting out can—and cannot—do

Artists who did not accept an advance can generally opt out of future participation. Artists who accepted an advance are subject to the required participation period discussed below.

Opting out can stop

• inclusion in newly created LANDR datasets;
• future licensing of the recording through the program;
• continued presence in certain time-limited datasets after a scheduled refresh;
• new training uses that have not yet occurred.

Opting out cannot necessarily reverse

• training already completed;
• existing AI models built using the recording;
• outputs already generated;
• rights already granted under completed partner arrangements;
• a copy already supplied under a perpetual dataset licence.

The advance can keep a catalog opted in for up to two years

LANDR’s terms establish a required participation period for artists who accept an upfront recoupable advance.

The participating recordings remain in the program until the earliest of:

1. the advance being fully recouped;
2. two years passing from acceptance;
3. the artist repaying the remaining unrecouped balance.

An opt-out request submitted during that period does not necessarily take effect immediately. LANDR says the artist can contact support, repay the remaining balance, and then leave early under the applicable process.

This does not mean the artist loses the advance or that LANDR automatically pursues an unpaid debt. It means the artist has exchanged immediate money for a temporary commitment to remain available for qualifying Fair Trade AI uses.

How much could an artist realistically earn?

There is not enough public payment history to provide a reliable typical-earnings estimate.

Long-term income may depend on:

• how much eligible audio is approved;
• how many datasets include it;
• how much each AI partner pays;
• the governing 20% or 25% rate;
• whether allocation is based on tracks, duration or another method;
• the deductions used to calculate net revenue;
• whether partner deals renew;
• whether a model or AI product earns meaningful revenue;
• how revenue is reported and audited.

The advance is measurable. The future licensing value is not yet measurable from the public information available.

Is $5 per track fair?

The answer depends on what the artist values, what the catalog is doing today and how the creator views participation in generative AI.

Five dollars is not necessarily LANDR’s claimed full value for a song. It is an advance against the artist’s future participation. The real evaluation must include the licence, the allocation method, the revenue reporting and the permanence of completed training.

Does the system reward quality—or catalog volume?

LANDR’s current terms describe pro-rata allocation according to the total seconds of qualifying audio contributed. That is more nuanced than simply counting tracks, but it still primarily rewards quantity or duration rather than publicly disclosed measures of creative or training value.

It is not clear from the public terms whether a recording receives more value because it contains:

• a rare instrument;
• an underrepresented genre;
• exceptional vocal performances;
• high-quality isolated stems;
• unusual rhythmic or harmonic information;
• clean, accurate metadata;
• language or cultural material scarce in other datasets.

A duration-based system is straightforward to administer. It may not reflect the relative usefulness of different recordings to a specialized AI model.

Should an artist switch distributors for this offer?

Do not migrate an established catalog solely because the deadline and $5 headline create urgency.

A distribution move should be evaluated across the entire business relationship:

• subscription and maintenance costs;
• commission and payout policies;
• store coverage;
• Content ID;
• collaborator splits;
• support quality;
• release migration and metadata continuity;
• ISRC and UPC handling;
• playlist and stream continuity;
• takedown processes;
• tax reporting;
• Fair Trade AI eligibility and expected value.

LANDR also says that leaving distributed music in stores without an active subscription can trigger a maintenance fee equal to the greater of 15% of monthly distribution income or $4 per month. That cost should be included in any catalog-migration decision.

For broader release-platform considerations, see the Jack Righteous AI music distribution policy guide.

Why LANDR is doing this

LANDR has traditionally operated across mastering, production tools, collaboration and distribution. Fair Trade AI extends that position into data licensing.

The likely strategic value is substantial. A rights-verified opt-in catalog can allow LANDR to become:

• a catalog aggregator;
• a rights-verification layer;
• a licensing administrator;
• a supplier of organized training datasets;
• an intermediary between independent artists and AI companies;
• a developer or partner in commercial AI music products.

This is an analysis of the business model, not a claim about LANDR’s private motives. The basic strategy is visible: distribution is becoming a way to monetize music not only through listening platforms, but through the technology trained on the catalog.

Why AI companies want rights-verified datasets

Building an AI system from random internet files creates serious legal and commercial risk. A developer may not know:

• who owns each recording;
• whether publishing is cleared;
• whether performers consented;
• whether the file contains an uncleared sample;
• whether the metadata is accurate;
• whether the company can prove permission.

A LANDR dataset can offer affirmative opt-in, standardized terms, rights warranties, centralized payments and organized metadata. That can make the dataset more valuable to a company trying to build a commercially defensible AI product.

The value proposition is therefore not only the sound. It is the documented chain of permission.

Why this matters beyond LANDR

Even artists who never participate should pay attention because this program can influence the emerging market price and structure for AI-training rights.

Future distribution platforms may introduce:

• separate AI-training controls;
• catalog advances;
• voice-model licences;
• stem and multitrack licensing;
• genre-specific dataset offers;
• premium rates for rare musical material;
• time-limited-only licensing choices;
• partner-by-partner artist approval;
• provenance and usage reporting;
• revenue participation tied to product success.

The larger industry question is whether creators will negotiate meaningful participation—or whether inexpensive early offers will establish a low market price before musicians fully understand the value and permanence of their contributions.

What a genuinely artist-friendly program should provide

1. A single, contractually confirmed revenue percentage.
2. Clear gross and net revenue reporting.
3. Disclosure of each AI partner receiving the music.
4. Artist approval before perpetual dataset licensing.
5. Separate opt-ins for training, voice, stems, lyrics and other uses.
6. Higher rates for specialized or scarce musical material.
7. Collaborator and split-payment support.
8. Clear guidance for PRO, CMO and publisher relationships.
9. Output-similarity safeguards and a dispute process.
10. Audit rights or independent reporting.
11. Partner-compliance and security requirements.
12. Strong voice, identity and likeness protections.
13. A meaningful share when downstream products become commercially successful.

LANDR’s program does not need to include every ideal feature before it can be considered progress. These standards show the distance between a consent-based starting point and a mature, genuinely transparent licensing market.

Questions to ask before opting in

Eligibility and ownership

1. Does my SOCAN, ASCAP, BMI or other affiliation make the recording ineligible?
2. Can a co-written song qualify when every writer approves?
3. How are producer, performer and neighbouring rights handled?
4. Are royalty-free loops permitted?
5. Are any sample-based recordings accepted?
6. What ownership documents are required?
7. Will LANDR explain why a submitted recording was rejected?

Licensing and control

8. Which companies can receive my recordings?
9. Will I be notified when my music enters a specific dataset?
10. Can I reject perpetual licences?
11. Can I approve or reject individual partners?
12. Can I restrict my music to defined AI uses?
13. Can my vocals be used for singing or voice-model development?
14. Can partners modify, remix or create derivative works?
15. What happens to the dataset if an AI partner is acquired?

Compensation

16. Is the governing share 25% or 20%?
17. Is allocation based on tracks, total seconds or another formula?
18. What expenses are deducted before net revenue is calculated?
19. Will artists see the gross value of each agreement?
20. How large is each dataset?
21. Do specialized recordings receive higher compensation?
22. Does the pool include revenue from AI products and outputs?
23. How are perpetual deals valued compared with recurring licences?
24. What has a typical participating artist earned to date?

Opting out

25. Has my music entered a perpetual dataset?
26. When is the next refresh for each time-limited dataset?
27. How much of my advance remains unrecouped?
28. How quickly does opt-out take effect after repayment?
29. What documentation confirms removal from future datasets?
30. What happens if an AI partner violates LANDR’s licence?

Creator decision framework

Jack Righteous assessment

LANDR deserves credit for recognizing three principles that the AI music industry has too often avoided:

• artists should be asked;
• training material has value;
• rightsholders should share in that value.

That makes Fair Trade AI more defensible in principle than obtaining music without affirmative permission.

But creators should not evaluate this offer from the $5 headline alone.

The actual decision involves a recoupable advance, strict ownership warranties, a broad worldwide licence, sublicensing to AI partners, possible perpetual datasets, downstream commercial outputs, limits on the practical effect of opting out, and unresolved inconsistency between the promoted 25% share and the 20% stated in the current terms.

The program may suit fully independent musicians who consciously support licensed AI development and are comfortable treating their catalog as training infrastructure. It is not an automatic yes for every eligible artist, and it should never be accepted without reviewing the exact agreement shown at enrollment.

Final thought

The most important part of LANDR’s announcement is not the five-dollar payment.

It is the recognition that music used to train commercial AI should be identified, licensed, consented to, documented and compensated.

That is a meaningful step.

Artists should still understand that a small advance can authorize uses with long-lasting consequences. The future value of a catalog may be much larger than the immediate payment, and completed machine learning cannot always be reversed by changing a setting later.

LANDR is offering independent artists a seat at the AI-training table. The remaining question is whether the information, control and compensation attached to that seat are strong enough.

Five dollars may start the conversation. It should not end the evaluation.

Continue your rights research

• AI Music Rights & Ownership Explained
• The Truth About Owning AI Music
• AI Music Rights in 2026: Suno Creator Checklist
• TuneCore AI Music Policy 2026
• The Suno AI Controversy Explained

Primary sources

• LANDR Fair Trade AI program page and FAQ
• LANDR Terms of Service

Educational notice: This article provides creator-business education and a plain-language reading of publicly available information. It is not legal, financial or tax advice. Program terms, promotions, deadlines and eligibility rules can change. Review the exact agreement presented to you and obtain professional advice when necessary.