The topic this year is ‘Artificial Intelligence and Peace’. This is part of paragraph no. 5 of the message: “In the future, the reliability of an applicant for a mortgage, the suitability of an individual for a job, the possibility of recidivism on the part of a convicted person, or the right to receive political asylum or social assistance could be determined by artificial intelligence systems. The lack of different levels of mediation that these systems introduce is particularly exposed to forms of bias and discrimination: systemic errors can easily multiply, producing not only injustices in individual cases but also, due to the domino effect, real forms of social inequality. At times too, forms of artificial intelligence seem capable of influencing individuals’ decisions by operating through pre-determined options associated with stimuli and dissuasions, or by operating through a system of regulating people’s choices based on information design. These forms of manipulation or social control require careful attention and oversight, and imply a clear legal responsibility on the part of their producers, their deployers, and government authorities.
Reliance on automatic processes that categorize individuals, for instance, by the pervasive use of surveillance or the adoption of social credit systems, could likewise have profound repercussions on the social fabric by establishing a ranking among citizens. These artificial processes of categorization could lead also to power conflicts, since they concern not only virtual users but real people. Fundamental respect for human dignity demands that we refuse to allow the uniqueness of the person to be identified with a set of data. Algorithms must not be allowed to determine how we understand human rights, to set aside the essential human values of compassion, mercy and forgiveness, or to eliminate the possibility of an individual changing and leaving his or her past behind.”