The broad principles of the ethics of artificial intelligence are gradually taking shape, but the question remains: who will ensure adherence to these principles?
The ethics group of the Artificial Intelligence Programme assumed an active approach to this issue. Rather than stopping at the definition of ethical principles, we went further to discuss the methods and competences required for the practical implementation of the principles.
In Europe, practical implementation began towards the year-end. Whether trust is an important element of ethical AI is no longer a relevant question; instead, we are focusing on identifying the right solutions and choices for building and maintaining trust.
This year we will see practical implementations of the ethical issues of AI, both in private companies and in public organisations.
New responsibilities call for new specialists
What can we achieve with co-creation and workshops? Do we need ethics specialists? These are frequently asked questions in many organisations. How do we define the responsibility of developers who build systems that may be incomprehensible to the administrative personnel?
Multidisciplinary teams could be the answer. People with different backgrounds look at ethical issues from different perspectives. Collaboration will bridge the cultural and ideological gaps, and help to create a working system. Involving everyone in AI development not only ensures ethical applications; it ensures the maximum distribution of the benefits of AI.
Specialists in ethics may be invited to support the work of development teams. A small but growing group of designers, scientists and software requirements engineers with in-depth knowledge of AI and ethics can provide systematic expertise. We need a wide range of skills and seamless cooperation in order to resolve the technical and non-technical challenges of AI ethics.
For AI developers, knowledge of ethics is becoming a new area of professional competence. An open mind towards other specialists representing different viewpoints is a key requirement. We need developers with knowledge and appropriate tools to ensure applications do not use discriminating data, threaten privacy, or reduce the decision-making powers of individuals.
Active dialogue between companies, NGOs and universities
Research into the ethics of artificial intelligence and public debate on the subject must continue on various forums. Fortunately we now understand that answers to many questions regarding the ethics of AI can be found from the tradition of ethics and professional ethics. This has facilitated the formulation of AI ethics related guidelines. At the same time, however, we need to bear in mind that we have only seen the first steps of artificial intelligence in its different forms.
Public debate on the ethics of AI must have an effect on the technologies being developed. The active discussion between the public administration sector, companies, legislators and universities seen in recent years is an excellent start. Hopefully this dialogue will result in practical cooperation. Close research and development cooperation provides a significant opportunity for Finland and will support the adoption of artificial intelligence in various organisations.