The balance between innovation and human rights: problems of applying artificial intelligence

Анотація

The article provides a comprehensive analysis of the impact of artificial intelligence technologies on the human rights system in the context of the digital transformation of society. The main threats and challenges posed by AI systems for the implementation of fundamental human rights and freedoms are investigated. Particular attention is paid to the problems of algorithmic bias, which leads to a violation of the right to non-discrimination, mass collection and processing of personal data without proper control, which threatens the right to privacy, restrictions on freedom of expression through automated content moderation, as well as threats to the right to a fair trial in the case of using automated decision-making systems without proper transparency.

The specific risks associated with mass surveillance and biometric identification technologies, including facial recognition systems in public places, the use of AI in employment and the military, and the manipulation of public opinion through deepfake technologies, are analyzed. Three stages of assessing the impact of AI on human rights are considered: analysis of the quality of training data, risk assessment at the system design stage, and consideration of algorithmic interactions. It is argued that AI systems, by their nature, reproduce social biases embedded in past experience data and do not have the inherent ability to change their behavior in accordance with the evolution of ethical norms in society. The application of artificial intelligence in the financial sector is examined in detail, in particular in credit scoring systems, where algorithms analyze huge amounts of data about the applicant’s digital footprint. The problem of “network discrimination” is identified, when a person’s financial capabilities are assessed based on the characteristics of their social environment, which violates the principle of individual responsibility and can limit freedom of belief through self-censorship. The example of the practice of American companies shows how the use of AI systems in financial decision- making can both expand access to credit for representatives of marginalized communities and strengthen existing forms of discrimination.