This guest post was authored by Ishita Khanna.
Information privacy theorists have argued that data is ‘quasi currency’ in the age of information technology. Though the economic value of data incentivizes companies to collect data, big data also possesses the potential to increase productivity, improve governance and thus benefit consumers and citizens. Data-driven insights can extract significant value by analysing it for purposes such as cost savings, enhanced procedures, a better knowledge of behaviour, and highly tailored products.
In this data driven economy, the data generated and collected by machines and human beings possesses tremendous value. Machine generated data is said to be data which is created through the use of computer applications, processes and services or through sensors which process information which they get from software, equipment or machinery, whether real or virtual. An interesting example of machine generated data is provided by the automobile industry. Sensors installed on cars generate data with respect to traffic prediction, location based searches, safety warnings, autonomous driving, and entertainment services. An analysis of this data can result in monetizable insights in the form of revelations for vehicle designs or selling access to the insurance industry. In this way, data precedes information, which precedes knowledge, which precedes understanding. Thus, raw data in crucial in the generation of value.
It was in response to this phenomenon that the European Commission in 2017 had proposed the creation of a ‘data producer’s right’ (DPR) which would protect anonymized, non-personal machine generated industrial data against the world i.e., a novel property right in data. This would create in favour of the data producer, ‘a right to use and authorize the use of machine generated data’. One of the other dominant reasons inspiring the call for creation of a novel property right in data stems from the fear that American companies are misappropriating valuable European assets. The use of European news by Google led to an initiative for a neighbouring rights for news publishers in the EU which furthered the call for a data producer’s right. For instance, the introduction of the sui generis database producer’s right in Europe in 1996 was borne out of the fear of domination by the US database industry over Europe’s markets.
This post seeks to critically examines the background, stated aims, subject matter and scope of the data producer’s right. It studies the inter-relationship between the existing intellectual property regimes and the property right in data to analyse if and how this new right would affect these regimes. Towards the end it offers recommendations for the alternative models that could be adopted for the protection of non-personal data.
Do we need a new IP right for machine generated data?
It has been contended that the existing IPR regimes as well as civil law, contract law and trade secret protection do not offer requisite protection to machine generated non-personal data since they do not create an ex-ante right in rem, hence raw data would not be protected from misappropriation by third parties and a market for licensing of data would not emerge. Copyright law only protects acts of authorship or compilations of data that are a consequence of creative arrangement or selection. Further, the sui generis database right only extends to data structured in a database. Hence, the argument for the introduction of a right in machine generated raw data.
The EU- DPR was envisaged as a novel type of intellectual property right, as the means to an end: to make data accessible. However, building new property fences seems paradoxical to the idea of increasing access to data. The answer to this lies in recognition of the fact that ‘property is an institution for organising the use of resources in society’. The stable legal entitlements that come along with a property right incentivizes the development of a valuable resource through consolidating both risks and benefits in right-holders and also stimulates the use and trade of data. The DPR was conceived of as a right in rem i.e. ‘enforceable against the world independent of contractual relations’ including the exclusive right in the data producer to use certain types of data and license their usage, thus embodying the essential features of the right of ownership of property. The hope was that infusing machine generated non-personal data with property rights, it would lead to the creation of a stable and safe licensing marketplace for the data.
Why is data such a challenging subject for IP Law?
However, the DPR would extensively overlap with copyright and sui generis database rights in production made using digital machines which could give lead to numerous competing ownership claims. For instance, the aggregate stock market data in a financial database would be the subject matter of protection of both the data producer’s right and the sui generis database right. Further, DPR could trump the statutory limitations laid down under the existing IPR regimes and the database right thus limiting their scope of protection. At present, users in the European Union are allowed to copy data from databases for the purpose of non-commercial research. The DPR would infringe on such freedoms allowed to the users unless it includes within its ambit all such relevant exceptions.
Another objection against a property right in data lies in its inherent lack of legal certainty and stability with respect to its scope, subject matter, and ownership, essential for it to be considered as a full-fledged IP right enforceable against the world. A property right in data would severely infringe on the freedom of expression and information by curtailing access to data to scientists, research institutions and journalists with respect to text and data mining. This freedom has been acknowledged in Article 13 of the EU Charter which stresses on the free flow of data in arts, scientific research and academic freedom.
Thus, a data producer’s right would encroach upon the central tenet of the IPR system which regards data as ‘free air for common use’ and only offers protection to creative and innovative inventions. The dynamic and fluid nature of raw data makes it difficult to classify as subject matter of a full-fledged intellectual property right. The database right raised a similar objection. However, the definition of ‘database’ and the requirement of a certain threshold of investment created at least some stability in the scope and subject matter of the right, unlike the DPR.
It is also important to understand why the property logic for data protection failed. The lack of success of the closely analogous, sui generis database right in promoting investment in and incentivizing the formulation of databases in the EU database industry is one of the reasons. Another reason is said to be attributed to the inclination towards opening data or making it accessible for both commercial and non-commercial re-use thus doing away with the exclusivity requirement. Hence, currently there exist no potent economic justifications for creation of a DPR. Instead, data producers can protect their data using the contract law, trade secret law and technology law protection mechanisms.
Thus, it can be concluded that a novel IP right should only be introduced after thorough economic-evidence based research establishing the real requirement for the right and not spontaneously. However, this alone will not suffice and must be accompanied with a methodical legal analysis of the scope and subject matter of the new right as well as its inter-relationship with the existing IPR regime.