Ethical Guidelines for AI in Public Administration

Responsible use of artificial intelligence requires good guidelines

Ethical Guidelines for AI in Public Administration

Public administration operates using public funds and with the consent and for the benefit of citizens. This legal and ethical relationship between citizens, government and public officials is also the foundation for the use of AI.

AI and smart technologies in general are tools and instruments for public administration and official activities. The use of these technologies is based on the general foundation formed by the values and norms of public administration and on the legislation regulating official activities.

Government authorities exercise of public powers and, when doing so, have a responsibility to act impartially, independently and equally. The activities of authorities must be open, appropriate and acceptable. Furthermore, authorities must ensure that the parties to a matter enjoy due process of law and that liability of public officials for acts in office is property realised. These requirements also apply to the use of AI in all official or administrative activities.

Government authorities are always responsible for the AI systems they use. AI is a support technology that can enable the use of various information technology devices and solutions in public administration. Though AI technologies are not independent actors comparable to humans, they are complex programs that learn more independently.

AI technologies have the capacity to modify and reorganise administrative structures and practices. AI technologies do not call existing values or norms of good governance into question, but can bring them to the fore in new contexts. AI technologies may have direct or indirect effects on offi-cial activities that previous information technology tools did not have.

The use of AI technologies requires that public administration is sensitive to the effects that AI has on citizens, the rights of citizens, society and the environment. Government authorities must also ensure that AI technologies promote the general objectives of public administration, such as administration that supports open, equal and active citizenship, the efficiency and appropriateness of administration, and a society based on trust and respect for fundamental rights.