Responding to Cambridge Analytica – The Role of Libraries

Recent stories about the activities – the misdeeds – of Cambridge Analytica have provided an illustration of the power of information.

It is true that digital technologies and techniques allowing the ‘mining’ of data have opened up unprecedented opportunities to gain understanding of our environment, and ourselves. Just as it has turbo-charged scientific research, it has also benefited market research, thanks to new possibilities to collect and use information.

It is a further reminder that technology is usually neutral in and of itself – it is what we do with it that defines whether its impact is positive or negative. We need mature approaches. Libraries, given their long experience in working with information and helping users make it work for them, have a lot to contribute.

Sadly, it has also provided an excuse to bring out old and lazy arguments implying that we should turn back the clock on technological change, or at least stop it,  regardless of the costs to innovation and progress.

In the case of an event at the European Parliament this week, Cambridge Analytica was brought out as a reason to make text and data mining activities conditional on using publisher APIs and subject to additional payments. The argument goes that text and data mining is dangerous, and so should remain under the control or supervision of rightholders, even where a user already has access.

However, this position is both paternalistic and regressive.

Paternalistic in that it presumes that rightholders should be able to supervise what analysis users are carrying out legally. And regressive in that it makes the ability to undertake research dependent on the ability to pay additional fees, over and above the initial costs of access.

Cambridge Analytica, in the present case, seems to have benefitted from data collection rules that were both loosely drafted and loosely applied. But with a business model based on the collection and use of data, and provision of services for wealthy clients, the company would be first in line to make the necessary additional payments under the model proposed in the European Parliament event. In this, they would be joined by other major ICT or pharmaceutical companies with the bank balances to support this. And in the case of Cambridge Analytica, the primary beneficiary of any additional payments would likely have been Facebook.

Such a step would exclude smaller firms and university spin-offs, as well as data journalists and individuals. A good result in the short term for publishing companies (and Facebook), but highly negative for innovation and progress.

A mature response to the issues raised by Cambridge Analytica would include the following:

  1. Privacy rules must be respected. This applies as much in the use of social media platforms as in publishing platforms where the collection of data about who and how is reading works is an increasingly central piece of business models in the sector.
  2. Users must be able to take conscious and reasoned decisions about what information they want to share. This requires education about how the Internet works, not just in school but also throughout life.
  3. Clear permission for anyone with legitimate access to data (respecting privacy laws, and with payment where appropriate) to carry out text and data mining, in order both to advance research, and permit the sort of data journalism that has allowed other recent scandals to be revealed.

Libraries have a role in all of these. In negotiating contracts with service suppliers, they can be attentive to privacy implications, and ensure maximum protection of users’ data. Libraries are already active in many places in teaching users about the operation of the Internet, and how to keep personal information safe. And they are currently arguing, in Europe and elsewhere, for a simple exception to copyright for text and data mining that will allow new ideas and innovations to rise up.

Read IFLA’s statement on privacy in the library environment,  and its position on text and data mining in the draft Directive on Copyright in the Digital Single Market. You can follow what IFLA and its partners are doing on the EU reforms here.