Jacques, Garstkam, Hvidd and Street (2018)
Contents
Source Details
Jacques, Garstkam, Hvidd and Street (2018) | |
Title: | The Impact on Cultural Diversity of Automated Anti-Piracy Systems as Copyright Enforcement Mechanisms: An Empirical Study of Youtube’s Content ID Digital Fingerprinting Technology |
Author(s): | Jacques, S., Garstka, K., Hviid, M., Street, J. |
Year: | 2017 |
Citation: | Jacques, S., Garstka, K., Hviid, M. & Street, J. (2017). THE IMPACT ON CULTURAL DIVERSITY OF AUTOMATED ANTI-PIRACY SYSTEMS AS COPYRIGHT ENFORCEMENT MECHANISMS: AN EMPIRICAL STUDY OF YOUTUBE’S CONTENT ID DIGITAL FINGERPRINTING TECHNOLOGY. 10.13140/RG.2.2.14443.54560. |
Link(s): | Open Access |
Key Related Studies: | |
Discipline: | |
Linked by: |
About the Data | |
Data Description: | The study uses an existing dataset developed by Erickson et al. in 2013, comprising 1845 parody videos (derived from 343 pop music videos) obtained from YouTube. The present study observes the takedown of these videos at intervals until 2016, determining whether manual takedowns or automated Content ID takedowns were employed following a copyright claim. Thereafter, the Herfindahl Index is utilised to measure the levels of cultural diversity after the removal of each type of takedown (with the aim of isolating the effects of automated takedowns). In measuring cultural diversity, the study considers not only availability factors, but also numbers of views, and characteristics such as the harshness of a parody’s critique. |
Data Type: | Primary and Secondary data |
Secondary Data Sources: | |
Data Collection Methods: | |
Data Analysis Methods: | |
Industry(ies): | |
Country(ies): | |
Cross Country Study?: | No |
Comparative Study?: | No |
Literature review?: | No |
Government or policy study?: | No |
Time Period(s) of Collection: |
|
Funder(s): |
|
Abstract
“Freedom of expression is closely intertwined with the preservation and promotion of cultural diversity. Simultaneously, protection of freedom of expression is promoted as an essential foundation for democracy in a pluralistic society. Although the aim of fostering cultural diversity is widely recognised, challenges remain in analysing the relationship and interaction between cultural diversity and freedom of expression. These challenges are likely to loom large in the digital environment.
Digital technologies are changing the way cultural works are being produced, distributed and accessed, but digital technologies are also changing the way online uses of cultural work are monitored. Under the auspices of the Digital Single Market Strategy, the EU Commission has proposed a new copyright directive and intellectual property enforcement strategies. In the mind of the Commission, achieving this aim requires video-sharing platforms to embrace sophisticated technological tools, termed in this paper as ‘automated anti-piracy systems’ (‘AAPSs’). AAPSs are capable of recognising content which right-holders have already identified as their own, and respond based upon standing instructions from the holder whether the platform should permit use or block it. The Commission is intending to endorse and render mandatory the - so far - unregulated and informal notice and take-down procedure established by private entities.
There are inherent risks in this privatisation of decision-making when digital content is blocked, filtered or removed. There is a lack of transparency in the design, implementation and use of AAPSs. This exacerbates the lack of accountability of intermediaries, and concerns that they fail to respect fundamental rights compared to in cases where the judiciary would be involved in the decision-making process.
Consequently, one may begin to wonder whether an obligation to deploy automated anti-piracy systems on online-sharing platforms might run counter to promotion of cultural diversity. By requiring private entities to determine what unauthorised use is an ‘infringement’, will there be enough (or any) breathing space for cultural diversity to thrive? After all, the digital environment enables distribution and communication of artistic and cultural expression across borders and cultures in an unprecedented manner. While many may have very little economic value, they may be, nevertheless, indispensable for individual fulfillment and participation in a democratic society.
Instead of considering whether the use of automated anti-piracy enforcement mechanisms is likely to violate international human rights law, this article looks at this issue from another perspective, and asks whether use of AAPSs is likely to impact on cultural diversity. By promoting cultural diversity as desirable ‘ends’, freedom of expression and copyright become the ‘means’ for enhancing the interaction between cultures and thereby promotes democratic values. In this article, we discuss our empirical research which sought to assess the diversity in cultural expressions featured on video-sharing platforms.”
Main Results of the Study
Summary statistics reveal that the number of takedowns between 2013 and 2016 doubled, resulting in 38.5% of the initial dataset being removed. Of this 38.5%, automated Content ID mechanisms accounted for 32.1% (e.g. 83.4% of all takedowns). The study finds that parodies which directly copy audio or visual aspects of copyright protected works are significantly more likely to be automatically blocked. The authors partly attribute this to the base algorithm of Content ID, which relies on matching portions of audio or visual work.There is a limited correlation between Content ID takedowns and cultural diversity. Some automated removals may in fact reinstate diversity; for example, videos from the USA and UK were overrepresented in the initial dataset, but Content ID also subsequently disproportionately removed videos from these countries (thereby making other countries more proportionately represented).“Weapon” and “target” parodies are more likely to remain, however a disproportionate share of highly critical parodies were removed.
Policy Implications as Stated By Author
In regards to parodies:The study raises concerns that automated technologies will frustrate the purposes of the introduction of a parody exception, noting that the base functionality of the algorithm cannot interpret true parodies (e.g. by relying on audio matches and not accounting for changes in lyrics). In regards to broader copyright policy:The authors recommend that legislation should actively promote cultural diversity online, potentially by expanding exceptions and preserving the increasingly shrinking public domain. Caution is noted against relying on private commercial entities to enforce copyright law, and instead an independent dispute resolution system could be introduced for takedown complaints. Rightsholders and intermediaries may wish to negotiate best practices for platforms, which may include providing better upload information for users, or revising “match” policies. Lastly, the authors suggest integrating a mechanism within platforms which allows users to directly obtain a licence to use a work.
Coverage of Study
Datasets
{{{Dataset}}}