Prior to the meeting, Christoph, Markus, Tom and Michele met to discuss the proposed changes to the data ethics codex. We have received 22 feedbacks, including 19 written feedbacks, 3 live interviews, and 18 feedbacks from companies within Alliance.
Our meeting was hosted by Peter Kolbe of SBB CFF FSS, in a building that is integrated within the structure of the Bern main station, where we could see trains moving on railways on the other side from our windows.
Peter Kolbe started the meeting by sharing with us some projects of SBB CFF FSS for using big data, the ethical issues they raise, and the procedures they set up to discuss these. Their ethics assessment workflow is (very roughly):
› 1. Describe the idea
› 2. Check the validity of the research questions (value for the company)
› 3. Privacy – legal check
› 4. Communication – check for added customer value, reputational consequences
We also discussed the question of an ethics of responsibility vs. a more utilitarian ethics and their impact on the companies’ ways of dealing with data.
We then presented a summary view of how we will modify the codex based on the feedback received.
Christian Hauser of HTW Chur presented his plan for a course to develop ethical motivation for companies. He asked the company representatives at the meeting to help him to identify those key figures in a company that could implement ethics if adequately motivated. The idea of this course will be developed as a project in the course of next year.
Another educational element was solicited by Karin Lange of Mobiliar: to develop case studies as teaching materials to explain the ethical codex in a practical way. It was discovered that some case studies have been developed already by Bruno Frischher of Luzern University of Applied science. Next year, Karin Lange, Michele Loi and Jean Gabriel Piguet will form a group to develop the case study further.
Beside the already mentioned educational projects, our plans for the next year are:
› Organizing and running the SDS conference in data scientist. We have already found a keynote for robust and trustworthy (ethical) AI, namely prof. Krishna P. Gummadi of the The Max Planck Institute For Software Systems of Saarbrücken
› Writing an implementation guide for our ethical codex, after running some implementation workshops with members of the Alliance, in the first quarter of 2020.
› Organizing an outreach event, which could be on the topic of the ethics of open data.
› Organizing an event on “how to set up a data ethics board” involving companies that have, or plan to build, one such ethics board.