Faculty

Claudia Roda

“We Need Ethical Approaches to Artificial Intelligence”

Claudia Roda is a true interdisciplinarian. She came to AUP in the 1990s to teach what was, at the time, a class in a young and innovative field: Introduction to the Internet and Web Authoring. Though employed as a professor in the Department of Computer Science, Mathematics and Environmental Sciences, she collaborates regularly with colleagues in fields as diverse as global communications and international law. Today, she is co-director of AUP’s new MSc in Human Rights and Data Science.

“I was coming with a research background in multiagent systems for large industrial applications,” explains Roda. A multiagent system is a set of artificial intelligence units that work together to be more flexible and achieve better results than the individual constituents would on their own. In the early nineties, Roda was part of ARCHON, one of the first EU-sponsored projects on AI. She later used this type of technology to research systems capable of adapting to users’ attentional needs. Her work with an interdisciplinary team was awarded a multimillion euro EU grant to study how, by tracking learners’ attention, educational material could be dynamically adapted to individuals’ learning needs. Intelligent agents tracked, among other things, users’ eye movements, keystrokes and responses to suggestions and prompts, with the aim of developing learning programs that would better hold users’ attention. Roda’s book Human Attention in Digital Environments (Cambridge University Press, 2011) outlines many of the results obtained in this project.

A problem emerged due to the scale of the personal data that was collected as part of the project. Given that the data included biometrics – and that her research subjects included children – there were ethical implications to consider. “We had incredible results measuring children’s educational progress in terms of metacognition,” she explains. “But there was this data protection contradiction that highlighted for me the human rights implications of this type of technology.”

Roda began researching ways in which engineers could design systems that both bring value and operate within an ethical human-rights framework. In collaboration with a team of researchers from industry and academia, including two AUP colleagues – Professor Susan Perry of the Department of History and Politics and Professor Antonio Kung from Roda’s own department – Roda founded the PRIPARE project, an EU-sponsored program aimed at preparing industries to apply privacy and security methodologies in the context of the General Data Protection Regulation (GDPR). “I started seeing all the ethical aspects that are really important to ensuring we develop advantageous systems that don’t detract from users’ personal freedoms,” she explains. “We need a different conversation between developers, designers and human-rights experts.”

The MSc in Human Rights and Data Science explores what form that conversation might take. Roda believes that consumer preferences will play an important role – similarly to how environmental concerns are now a large part of consumer choices. Her work through PRIPARE involved examining the full range of stakeholders involved in implementing data privacy regulations, highlighting where the knowledge gaps were, and then coordinating the design of educational materials to encourage ethical systems development. “We held workshops for different stakeholder groups, creating a curriculum that we then developed into a book,” she explains. Human Rights and Digital Technology: Digital Tightrope was published in 2017 by Palgrave Macmillan; the book in turn formed the basis of the idea for the new MSc.

Roda believes that, following the successful implementation of the GDPR, it is vital to develop a similar regulatory framework for data science in general and artificial intelligence in particular. “The ethical application of data science is a big problem,” she argues. “From issues of data bias to the transparency of the algorithms used, our challenge now is to find technical and regulatory solutions for the ethical use of AI.” The issue is that solutions need to be approached from a place of collaboration – policies implemented by individual actors are unlikely to be as effective. International organizations such as the European Parliament, OECD and UNESCO are a vital part of the conversation, and Roda is hopeful that their efforts will bear fruit. "I believe the master’s program is particularly exciting in this context,” she says. “By looking at both ethical and technical questions surrounding data science, our graduates will be able to take a proactive stand in the multistakeholder conversations necessary to build and regulate the type of technology we want for our future.”

---

The MSc in Human Rights and Data Sciences explores the ethical implications of AI alongside those of other digital technologies including Big Data and the Internet of Things. Roda brings her broad, interdisciplinary background in AI, attention and data privacy to her teaching, which, alongside the human rights expertise provided by Professor Susan Perry, provide a rigorous, forward-looking foundation for the program.