Velislava Delcheva warns of AI risks in decision-making

Bulgarian Ombudsman Velislava Delcheva has warned of the significant risks posed by the use of artificial intelligence in decision-making processes that affect citizens' lives. While acknowledging that AI can assist institutions and businesses, she noted that the primary concern arises when AI is used to make critical determinations regarding employment and the exercise of civil rights.

Delcheva highlighted that AI is increasingly being used in recruitment, including creating job advertisements, preparing tests, conducting interviews, and evaluating candidates. This raises urgent questions regarding accountability and the transparency of the criteria used to reject applicants.

A major risk identified is the potential for AI systems to perpetuate human-coded discrimination. For instance, if an algorithm is programmed with biased criteria, such as disadvantaging older candidates, it could systematically restrict access to employment. Delcheva emphasized that because the humans designing these systems set the underlying criteria, the AI can effectively enforce and automate discrimination.