DRAWING THE LINE BETWEEN DATA AND INFORMATION

INITIALLY

The Data Act came into force on the 12th of January, and although the transition period leaves plenty of room for discussion, many questions still linger around the Act and its implications. One of these questions has to do with a surprisingly far-reaching subject – The scope of the act regarding different categories of data. The ambiguity of the scope stems from the fact that the dividing line between different categories of data leaves plenty of room for interpretation – on paper and in practice.

The Act defines three types of data: Raw data, pre-processed data and derived or inferred information. The first two types are in scope of the act (and hence subject to data sharing obligations), while derived or inferred information is left outside the scope to protect the data holders’ business interests. Thus, the Act presents in essence a not-so-novel question many have sought to answer but none have exhaustively unravelled: What is the difference between data and information? The approach to such a philosophical question is best done in two parts, firstly through the frame of reference reflected as the relevant legal text on paper, and secondly through the effects and implications of the text in practice.

ON PAPER

According to Recital 15 of the Data Act, raw data is classified as data, which is not substantially modified, and is automatically generated without any further form of processing. As such, raw data would mean the first-hand data a connected device or service generates, a spreadsheet of illegible numbers and symbols. Pre-processed data on the other hand is data which is processed for the purpose of making it understandable and usable prior to subsequent processing and analysis. The Act strives to clarify the degree of processing relating to pre-processed data by determining that the processing should not involve significant investments to data curation and alteration. In other words, the data which results from the first processing activities is possibly still considered in scope of the regulation and thus subject to data sharing obligations, as long as the degree of processing is extended purely to the purpose of making the data usable. Finally, derived or inferred information is described as involving additional investments into assigning values or insights from the data, in particular by means of proprietary complex algorithms, whether standalone or a part of proprietary software. The notion of value should be understood as conveying the meaning of the data, rather than merely presenting it in a coherent manner.

IN PRACTICE

The line between pre-processed data and derived or inferred information is crucial for creating value from data, as it determines the spectrum of possibilities for organizations when it comes to business in the data market. With the above-mentioned frame of reference, it would seem that the key distinction between pre-processed data and derived or inferred information, or in other words in-scope and out-of-scope data, is the complexity of related data processing. This is because applying complex processing methods would automatically indicate significance of investments to the processing of data, as well as the potential for insights and values beyond readability and usability. This in turn might mean that for some organizations, their existing proprietary processing methods might be too complex for the purposes of generating pre-processed data, because most existing operations were likely built with the intention of deriving value from processing raw data and not just simply make it readable. Thus, organizations could be both overprepared and underprepared regarding data sharing obligations. Whichever the case may be, the need for implementing sufficient or ”non-significant” data processing mechanisms remains a key measure in Data Act compliance.

Herein might also lie the key to creating novel business solutions during the age of the Act: The sharing of pre-processed data in itself creates value, for data holders as they are provided under FRAND terms to data recipients, and for data recipients as the Act in essence legitimizes their market of operations. The providing of pre-processed data and data centric business in general also creates room for more refined and insightful services based on derived or inferred information, as these solutions could be seen as akin to ”higher tier” products in tier based business models. While it is evident that data sharing directly or indirectly based on a fee has been a way to conduct business long before the Data Act, the framework built by the Act will certainly drive innovation to generalize the offering of such products and services, as well as diversify the variety of products and services in the market.

Interestingly enough, the importance of discerning between in-scope and out-of-scope data seems to be more important in the context of related services. This is because the Data Act takes a very specific stance on the use of data when it is shared to competitors: A non-compete obligation to develop a product which competes with the product the data originated from. This phrasing clearly protects the physical devices instead of their related services, even if the device wouldn’t function without the related service. Alas, organizations seem to be afforded a narrower regulatory protection when it comes to related services they provide. This makes sense since the purpose of the Act is to facilitate data sharing and capitalize on its potential value on a large scale.

FINALLY

As the Act denies the development of manufacturer-competing products while staying silent about the development of competing related services, it is evident that organizations with related services as an essential part of their business stand to gain the most of discerning between what they are obliged to share and what is for them to derive value from. These organizations should therefore be as meticulous as possible when mapping the flow and degree of processing of their data, while ensuring they are not sharing data that is “too processed”, i.e. too valuable and thus out-of-scope of the Act by building different pipelines for different purpose data for example. These organizations should also carefully consider the specific terms they share their in-scope data with to maximize value gain. While these terms are subject to fair, reasonable and non-discriminatory (FRAND) conditions, a certain degree of manoeuvrability remains with the data holder. While no definitive or philosophical answers are to be found from these preparatory operations, when it comes to the Data Act, these measures take organizations considerably closer to drawing the line between data and information.

Want to learn more? Contact us:

Counsel Otto Lindholm, +358 50 378 7358

Associate Sohrab K?nk?nen, +358 50 432 1796