Response to the European Union call for evidence: Delegated Regulation on data access provided for in the Digital Services Act

  I am filling in this feedback in my capacity as a researcher. My work examines international security in digital spaces, especially social media platforms. There are, broadly, two ways researchers can obtain social media data: either collecting this by hand or using a programme to "scrape" it automatically from a given platform. I use the former, but a significant proportion of researchers use the latter, and I have familiarity with its methods. 

Twitter is, at present, attempting to prevent researchers from using the scraping method for free, as it wishes to monetise this element. The advantage of using a scraper is that you can access significantly more data, which better allows for quantitative and natural language processing analysis. Business can afford to pay the fees for this sort of data in order to obtain market advantages. Without the ability to access this method, researchers will find it much harder to conduct larger scale work especially. The kinds of work that might be affected include the ability to understand social media-based election interference, track conspiracy theories - which grow ever more prevalent - and monitor hate speech, which can be an indicator of potential political violence. An inability to access this sort of information is a significant loss to the public good and it must be protected in law. Technological developments exceed our current understanding of their operation socially, politically, economically and more. Despite this, I believe it is clear that private ownership of companies such as Twitter should not prevent some sort of regulation for the greater public good. 

In terms of the specific laws proposed, access to social media data would be granted to researchers and this is something I strongly support. There is, however, unevenness in ethical practice within social/digital media research methods. In my view, some practices are not as ethical as they could be, largely because digital data is perceived as either "public" or "not public." In reality, this is much more complex and the division is not a binary one. There needs to be a level of monitoring within the process of granting data access, to ensure that ethical practice best balances the public good of analysing digital data with the need to ensure that harms are minimised. This is, perhaps, more likely to be problematic in smaller studies, where individuals might be identifiable but is a concern that all researchers and regulators should take seriously.