Responsible AI: An Urgent Mandate
Ricardo Baeza-Yates,Usama M. Fayyad
DOI: https://doi.org/10.1109/mis.2023.3343488
IF: 6.744
2024-03-02
IEEE Intelligent Systems
Abstract:AI is rapidly becoming essential in various industries, raising societal expectations. AI's societal consequences include impacts on mental health; misinformation; workforce displacement; and economic, regulatory, and law enforcement challenges. Indeed, the regulation of AI usage is on the horizon, with the European Union and China already taking big steps, while the United States drafted its first AI-related bill of rights last year. Professional associations and other nonprofits are also contributing to AI ethics and regulations, increasing the urgency and criticality of this area. In this new context, public services and regulated institutions must ensure responsible AI to avoid biased or inaccurate decision-making. Similarly, companies using AI responsibly can stand out, increase efficiency, and avoid future legal problems. This article highlights the issues and problems that result in many organizations not knowing how to do responsible AI in practice, as they need to identify potential problems, set up safeguards, and conduct ethical impact assessments, among other actions. We present the issues to consider toward a comprehensive approach to responsible AI that should include defining a responsible AI strategy road map; assessing models, processes, and products; and training individuals at different levels. By covering the pressing issues related to the urgent need for adopting responsible AI, we hope to highlight the importance for corporations to seriously consider responsible AI as they rush to adopt this technology for competitive advantage.
computer science, artificial intelligence,engineering, electrical & electronic