On March 30-31, Chancellor of Justice Ülle Madise participated in a meeting at the invitation of the European Commissioner for Human Rights, Dunja Mijatović, where they discussed how to apply artificial intelligence in such a way that human rights are protected.
The Chancellor of Justice shared Estonia's experience and emphasized the need to distinguish automated administrative decisions from decisions made with the help of artificial intelligence, as well as the answer of a chatbot without binding consequences from a decision with binding ones. In the latter case - if they should be allowed in the relationship between a person and the state authority - it is important from which sources the machine obtains information about the person and whether this data is correct, according to which logic the data is analyzed and how the result is reached. What is important is that the person knows that the decision was made by a machine and that they can challenge it effectively if needed.
With the support of artificial intelligence, the protection of people's rights and freedoms can be significantly improved, because the machine is able to solve problems quickly and without mistakes, and does not commit a crime of corruption. Unfortunately, an uncontrolled machine can do bad things just as effectively: amplifying unequal treatment and past mistakes, for example. Therefore, Estonia also needs to move forward with Bürokrat and other artificial intelligence solutions, but it is necessary to regulate both automated decisions and the use of artificial intelligence. In any case, caution is in order.
In the photo, Chancellor of Justice Ülle Madise with her Dutch colleague Reinier van Zutphen and Ukrainian colleague Dmytro Lubinets.