CORPORATE GOVERNANCE OF RESPONSIBLE ARTIFICIAL INTELLIGENCE IN LEADING IT CORPORATIONS: A COMPARATIVE ANALYSIS OF PRINCIPLES, INTERNAL STANDARDS, AND PRACTICES

Abstract

This article presents a comparative analysis of corporate governance of responsible artificial intelligence in leading IT corporations (Google, Microsoft, IBM, Meta, Amazon, Apple, OpenAI) by examining declared principles, internal standards, and practices of their practical implementation. The study is based on a qualitative content analysis of publicly available corporate materials and follows a framework logic of “principles → corporate governance → processes → control mechanisms → monitoring → feedback.” A comparative matrix for assessing the maturity of practices is proposed, based on nine criteria (fairness, transparency/explainability, safety/robustness, privacy, accountability/human oversight, social benefit, corporate governance, monitoring/reporting, product restrictions) using a three-level scale (0–2). The findings show that despite a shared “value core” of responsible AI, the level of institutionalization and operationalization of principles varies significantly across companies and depends on product profiles and risk exposure. Several устойчиві patterns are identified: the transition to risk-based governance as an operational norm, the standardization of transparency artifacts (model/data documentation), and voluntary self-restrictions in high-risk areas as an indicator of accountability. The generalized framework for corporate implementation of responsible AI and the comparative results can be used to interpret the maturity of corporate models and to adapt best practices in organizations operating under increasing regulatory and societal demands.