[Guest post] One Database to Rule Them All? The Musical Works Database under the Music Modernization Act

The IPKat is pleased to host this guest contribution by Natasha Mangal (EIPIN Innovation Society; CEIPI, University of Strasbourg), which reflects on the recently adopted Music Modernization Act [Katposts here].

Here’s what Natasha writes:

Lawmakers in the U.S. recently passed the ambitious Music Modernization Act (MMA) promising, among other aims, to create a new compulsory blanket licensing system for mechanical rights in certain digital exploitations (online streaming and digital permanent downloads). Importantly, the Act has far-reaching effects not confined to U.S. musicians and works, but applies to all songs written that are exploited digitally in the U.S. market. Though there are many interesting aspects of the Act, this post will focus on Title I, and particularly its provisions for the development of a new, publicly-available database for musical works information.

Licensing musical rights has always been a complex endeavor. In a single piece of music, there are two underlying rights: one in the sound recording and one in the composition. In each of these, there is a “bundle” of exclusive rights: the right of reproduction, right of public performance, the right to prepare derivative works, and so on. There is a multitude of stakeholders that partake in the royalties generated from the exploitation of these rights: in a single piece of music, there can be as many as 15 parties with an ownership interest in the work. The complexity inherent in music rights ownership has been compounded in the digital age, as a lack of common metadata standards and distribution solutions have given way to often impractical and disjointed approaches to rights management. With thousands of new works being hosted online daily, the licensing problem has quickly escalated for all parties concerned. 

With regard to the compulsory statutory mechanical license for musical works in the U.S. (17 U.S.C. §115), the process for obtaining the license used to be on an individual, per-work, song-by-song basis. This led to millions of NOI filings on the Copyright Office by digital music providers who, at times, claimed they “could not locate” some famous songwriters. On the basis of this system, Spotify and other big media platforms were sued by rightholders and music publishers multiple times for failing to obtain a proper license for the music hosted on its services, an error that skyrocketed damages claims into the billions.

Yet the issue was not simply one of negligent filing: neither the Copyright Office, nor the various agencies (i.e., HFA, Music Reports), which were hired by the platforms to administer the licenses, had a complete picture of the necessary data to determine ownership.

The idea to create a single, centralized, authoritative database for musical works information is not a new one. One previous effort that came close to actualization was the “Global Repertoire Database” (“GRD”) project, which was started around 2008-09 between the collectives PRS for Music, STIM, SACEM, and involved several other stakeholders such as major music publishers and tech companies. The GRD project began with a strong foundation from a technological standpoint: the data standards and pooled multi-territory database of over 15 million works that were being developed by the International Copyright Enterprise (ICE) at the time, were to be used as an early basis for the database.

However, by 2014 the project ended in deadlock. One issue regarded ownership of the underlying data. Collecting societies had, until that point, maintained some level of proprietary interest in developing their own databases, and there was uncertainty regarding future ownership of the data. Another looming issue was the massive amount of funding required: by the project’s end, early estimates of the GRD’s operation costs were vastly exceeded, with concerns growing over the amount collecting societies would have to contribute to keep the project afloat.

How might the database effort under the MMA be different? First, digital music providers such as Spotify and Apple Music will be the sole parties responsible for funding the operations of the new Mechanical Licensing Collective (MLC), including the costs of building and maintaining the database. Second, the introduction of the MLC itself – a centralized, non-profit institutional actor – will ideally serve as a neutral party to facilitate administration of the license, and will perhaps provide a more centralized point of information and dispute resolution for interested parties. As for ownership of the underlying data, for now it seems that the “public” nature of the database will ensure that no one party can claim ownership. However, the availability of “bulk copies” of the database (as described below) to third parties might potentially become an issue later on.

There are many stakeholders interested in the process of determining what data standards will be used and how they will be integrated into the new database. Currently, there are many music metadata standards in use. Some standards, such as those approved by DDEX, are already familiar to music publishers and societies but are not explicitly mentioned in the Act, whereas other standards such as ISWC (International Standard Musical Work Code) and ISRC (International Standard Recording Code), are specifically mentioned.

Therefore, there seems to be some acknowledgement that there is no need to “reinvent the wheel” – what will matter is proper implementation of already existing data standards. One organization that will likely play a role in this adaptation phase is CISAC, an international umbrella organization of CMOs which pioneered the ISWC standard.

Accessibility of the data contained in the database, especially in terms of obtaining a so-called “bulk copy” of the database, is also an open question. For now, users of the blanket license will be able to obtain such a copy of the data, but it remains to be seen whether, and to what extent, other third parties (i.e. MLC vendors) may be able to do the same. While access itself may not be problematic, commercializing on the data and/or using the data to create competing databases in the market may only lead to further confusion and administrative fragmentation. Several stakeholders who contributed to the Copyright Office’s Notice of Inquiry also listed this as a concern.

Finally, if this database is indeed held out as “authoritative” in the music industry, it is questionable whether the data can then be used as a basis for litigation. This could become problematic if the data represented in the database is as full of errors as under the previous regime. To avoid this scenario, the database needs to be sufficiently up-to-date and the parties providing the data – for now, mostly the digital media platforms – need to use “good-faith, commercially reasonable efforts” to ensure data accuracy. Otherwise, the database must include a disclaimer on its authoritativeness, but this option would seem to undermine the overall mission of the Act.

It is foreseeable that databases like the one envisioned in the MMA can be used in the future for automizing certain processes in copyright licensing. As mentioned above, if third parties are able to obtain and use the publicly-available data in the database in bulk, they can potentially build upon that data with their own technology. An interesting parallel can be made with the publicly-available data provided by the National Weather Service (U.S. governmental agency) and third-party use of the weather data derived from the service. 

Perhaps it’s also worth noting that one vendor already designated by the MLC, ConsenSys, is a company that develops blockchain technology solutions. This might hint at even more future-looking copyright licensing practices that involve using the database information in innovative new ways.

The database is staked to become operational by January 1, 2021. For now, we will have to continue to watch closely for specific regulations that will adopted by the Copyright Office.