While AI technologies have number of positive effects on our lives, they also give rise to a plethora of ethical issues as the design and use of AI are inevitably socially and culturally embedded. For example, machine-learning methodologies can give rise to inadvertent biases in decision making, which derive from the programming choices and data used to train the systems. Such biases may relate e.g. to individuals’ gender and ethnic background and affect their equality of opportunity.
There are only few studies on practical implementation of ethical guidelines or governance experiments of AI in the varying use contexts. In the ETAIROS project, the target is to study and co-create governance frameworks and tools, which respond to the challenge of socially sustainable use of AI.
In order to reach the project goals, there is a need for a shared and coordinated understanding across administrational sectors on general principles how AI is implemented and regulated. This includes also coordination of actions with the private and third sector. In addition, several factors in AI implementation should be noticed. For instance, adequate and symmetric information between stakeholders, coherent policy coordination and visioning across policy sectors, and sufficient transparency of activities will enhance successful implementation.
The main research questions to be tackled are
- What are the future trajectories of AI and its societal impacts?
- What kind of action-based ethics and risk analytics could govern future AI?
- How can we model and design AI systems that are ethically acceptable and how can AI help govern its own ethics?
- What kind of governance frameworks, practices and competencies are needed to foster sustainable, risk-free use of AI in the society?
To answer the research questions, the project will a) create context-sensitive use scenarios for AI technology, b) provide robust model of the ethically critical features of AI-based processes and risks, c) yield design practices and tools for ethical AI in the context of governance, d) develop AI aided tools for screening ethical aspects in governance, and e) provide functional forms of governance for the use of AI.
The research project is implemented jointly by University of Tampere, University of Helsinki, University of Jyväskylä, University of Turku and VTT Technical Research Centre of Finland Ltd. 4FRONT Pvt. Ltd. is responsible for the communications and interaction.