Missing Royalties? How Independent Songwriters Can Recover What They Are Owed

Renato Horvath
6 minutes
2026-03-01

When we ask independent artists what help they actually want, nobody requests an AI that writes songs. The requests are much more specific. One artist in our survey wished that "when I submit the songs, most of the fields would be already automatically filled out according to previous submissions, and once submitted, the ISRC would be automatically generated". Another wanted distributors to work directly with authors' societies, "so when you put in the ISRC, all the info from a song appears automatically". A third asked for something even simpler: a chart showing which song is generating which money, and from which platforms.

Notice what these are. They are not creative tools. They are requests for plumbing, for the administrative side of music to stop demanding expertise nobody ever taught them. That is where AI is genuinely useful for rights holders, and it is what we are building at Artistic Futures.

Rights administration is a data problem

Royalties fail to reach their owners because information about who owns what is incomplete, inconsistent or formatted differently across systems that do not talk to each other. The same song can be registered under slightly different titles in three national databases. A songwriter's name can appear with and without diacritics. IPI numbers, the identifiers that connect songs to their registered owners, are often missing from release metadata entirely.

Fixing this manually means cross-checking dozens of collecting societies and platforms, in several languages and national formats. No individual artist can do that, and our survey shows most never try: when a problem comes up, the standard move is to email the distributor's support team and hope. But cross-referencing messy records across many sources is exactly the kind of pattern-matching work that machine learning does well. Software can scan rights databases, flag probable matches for a human to verify, and surface the gaps: registrations that seem to be missing, catalogue entries that do not reconcile, territories where a work earns revenue but no claim has ever been filed.

Why Europe makes this harder

Every EU member state has its own collecting society, its own registration requirements and its own rules for mechanical royalties, performance royalties and neighbouring rights. BIEM societies, which handle mechanical rights in most European countries, long operated under rules that prevented authors from collecting their full mechanical share through a publisher unless specific conditions were met. An independent songwriter working across borders either needs a well-resourced publishing deal or the knowledge to register with multiple societies alone. AI tooling can at least map that maze: which societies matter for a given catalogue, which registrations are missing, where the most likely money is.

What to expect, honestly

AI will not recover your royalties on its own. The databases collecting societies maintain, the metadata standards platforms enforce and the interoperability between national systems all still need to improve. A tool can spot that your catalogue exists under three name spellings across European databases; it cannot file the correction for you. Human action, often direct contact with a collecting society, remains the last step.

The realistic promise for the next few years is a different one: cutting the time and expertise needed to understand your rights situation from months to hours. Knowing what you own, where you are unregistered, and what you are probably owed. For artists doing all of this alone, that changes the economics of even trying.

Built around what artists said

Artistic Futures is a Creative Europe project developing exactly this kind of platform for independent European songwriters and composers, and we are designing it with artists rather than for them. One survey respondent asked for "more robust programs and tutorials on metadata so that artists can protect their work". That request, and dozens like it, are the actual specification we are working from. If you write or compose and want to follow the development, or add your own answers to the research, take the survey.

Subscribe to the monthly brief

Royalties go missing quietly. We're documenting how, and building something to stop it.

Subscribe to the monthly brief