We want to avoid a situation where the use of artificial intelligence brings threats to the people, the economy, democracy and the sustainability of the country. At the same time, we also want keep up with the technological progress in order to ensure the competitiveness of the economy and individuals. Risk analysis is an appropriate tool to balance these objectives. Methodical risk analysis helps to justify decisions and investments, thereby preserving and increasing society's confidence in the use of algorithmic systems.
The main task of the study is to support the Estonian state and institutions in risk management process. To analyse risks, it is necessary to be aware of the threat and understand the simpler attack vectors related to the threat. But just being aware of the danger does not help. A list of reasonable mitigation measures for each risk is needed. These can be implemented by a state or organisation in order to reduce the likelihood of a risk materialising.
Research questions: What threat events caused by artificial intelligence are realistic at the Estonian and European level? What are the sequences of events, assumptions that lead to these threat events? What mitigation measures are already in place today and what more needs to be implemented? What is the capacity of Estonian and European organisations to implement these mitigation measures?