Actualizado por última vez en
A lot of industries seem to be overwhelmed by the rise of artificial intelligence (AI). Some proponents dream of utopias, while some opponents have dystopian nightmares. There is a lack of consensus in interpretations of the current legal and regulatory environments covering the implementation of computer systems leveraging AI and the related concepts of machine learning and deep learning. Interpreting, implementing and complying with the regulatory environment is not possible without unpacking the foundations of automated decision making.
FABRICS is an unpacking of artificial intelligence and an investigation of the regulatory challenges to AI and artificial decision making, in particular the “right to explanation” arising from the EU General Regulation on Data Protection (GDPR). FABRICS unpacks the challenges of explaining automated decisions and points to the need for privacy-by-design in the development of tools that use AI.
FABRICS also tries to address some of the challenges of AI through design: a system of warning signs that inform people of the presence of AI is outlined as a basis for activism, as well as inspiration for privacy-by-design approaches to AI. It also includes a piece of design fiction, which helps unpack the challenges of processing personal data on the edge of cyberspace.
Regulatory considerations in artificial intelligence
Systems, code and law
The effects of automated decision-making
On privacy by design
Automated decision-making and the GDPR
The right not to be subject to automated decision-making
A right to an explanation?
The two types of interpretations of the right to explanation under the GDPR
Warning and empowerment
Unpacking system transparency
The way forward
About the authors:
Alex Comninos is an information and communications technology researcher and consultant from South Africa who has published research on various topics including human rights and the internet, internet governance, cybersecurity, intermediary liability, mobile banking, and African political economy. Alex is a member of the Association for Progressive Communications and has consulted and conducted research for the Open Technology Institute, Freedom House, the OSCE and the World Bank. Alex is a podcaster at CYBERnyama.net.
Martin Konzett is a leading software architect and engineer, designer, artist and lecturer from Austria with over 30 years of hands-on experience with code, and an over 20 year track record in successfully delivering software solutions with core competences in Industrial Software Engineering (SE) and Human Computer Interaction (HCI). Martin has a strong focus on project operations, project culture and peopleware. He is a founding member of the Austrian ICT4D chapter and has consulted for governmental, intergovernmental and corporate bodies.