| Yhteenveto: | This Master's thesis assesses how to extend a higher-level developmental method
for trustworthy artificial intelligent systems, ECCOLA, by evaluating it with
Information Governance principles. Artificial intelligent systems are ubiquitous,
with their application prevalent in virtually all sectors. In addition, Artificial
intelligent systems rely on data and information they collect from users for their
development. These issues have prompted ethical concerns, especially as their
usage crosses boundaries in sensitive areas such as health, transportation, and
security, calling for better governance. As such, there is a need for developing
ethical artificial intelligent systems with effective governance that users can
trust with their information. Several guidelines exist to help facilitate these developments;
however, very few transition into methods with virtually no method
existing for higher-level development methods. ECCOLA is proposed as a
solution in transitioning from guidelines to development methods at higher
levels. The study extends ECCOLA by evaluating its ethical tenets with Information
Governance principles (Generally Accepted Recordkeeping Principles,
GARP®) as a governance framework to improve its robustness in line with ethical
guidelines. This was accomplished by following the Design Science Research
methodology approach using a conceptual framework based on ethical
guidelines of the European Commission and content analysis. The findings reveal
a vulnerability of the GARP® principles of Retention and Disposition in
ECCOLA. A possible solution artifact has been developed, which remains to be
tested.
|
|---|