DataharvestEIJC2020

Proactively automating family benefits in Estonia. Detecting medical prescription fraud in Portugal. Assigning GCSE and A-level grades in the UK. Monitoring smiles in bank branches in Poland. Implementing face recognition technologies to apprehend criminals, enforce social distancing or even fight racism. And many, many more examples of automation in the distribution of rights and services across 16 European countries. Presentation of the 2nd Algorithmwatch report on automated decision making in Europe.

Algorithms rule our lives, some say. But how do we know? Many automated systems are hidden from public view due to "security", "privacy" or "intellectual property" concerns. Those that are visible are often black-boxes with no accountability. Based on a dozen investigations by AlgorithmWatch and others, this session will explain how to develop methods to spot algorithms, audit them and communicate your findings.

Donate your data, so journalists can do journalism for you to find out, what your data are used (or abused) for. That's the logic behind the two data donation tools presented today: Dataskop by Algorithmwatch and the PersonalData.io toolbox. Paul-Olivier Delhaye, the founder of Geneva based non-profit PersonalData.io has recently collaborated with Swiss daily newspaper Le Temps to help readers chase their "digital shadow" based upon their own data. From Germany, Lorenz Matzat, co-founder of Berlin based non-profit Algorithmwatch, presents the data donation tool Dataskop. Hear about the tools and the potential stories. For background links, do have a look at algo-watchers meeting point during Dataharvest Digital 2020.

Automated decisions are on the rise in Sweden's social welfare system,shifting contact between authorities and citizens in need from personal to digital solutions and potentially hiding official decisions in black boxes of secret algorithms. The change has met criticism and journalists and academics address the question. One of them is Anne Kaun from the Södertörn University College. Together with colleagues she has filed a massive freedom of information (FOI) request to 321 government agencies to get an overview whether they work with automated decisions and in which areas they use it. Further, journalists and academics have filed FOI requests to gain access to the algorithms. Some FOI requests had to go to court to get the algorithm as the software provider claimed business secrets. In this session, Anne Kaun will present the work with    defining, tracking, analysing and contextualising these automatedprocesses as academics and journalists do in Sweden.

Does Instagram's algorithm promote nudity? Anecdotal experience by instagrammers could indicate so, but how to find out? Algorithmwatch journalists decided to set up a data donation option and analyse the data they got. Included were also other informations such as patent description of the inner working of Instagram. Come and listen to the "how-to" of this algorithm investigation. https://algorithmwatch.org/en/story/instagram-algorithm-nudity/

Networking round table: Are you algowatching? Do you want to start algowatching in your country? Share your story, meet experienced colleagues, get collaborations started.