The AI for Good conference organized by the International Telecommunication Union (ITU) in Geneva, Switzerland, this year marks a decade of global efforts to steer artificial intelligence (AI) for the well-being of humanity. However, behind this noble ambition, the event revealed a contrasting reality: the fast-paced development of AI technology often loses its moral compass due to the lack of a universal agreement on what is meant by "good."
ITU Secretary-General Doreen Bogdan-Martin emphasized that AI's potential to address global crises such as hunger and climate change is immense. Nevertheless, various parties have begun questioning the dominant role of big tech corporations. Sharp criticism has emerged from activists and academics who highlight the lack of transparency and the public sector's dependence on private entities, which risks widening the digital divide on a global scale.
Practical engineering perspectives further complicate the situation. Engineers report difficulties in translating abstract ethical concepts into concrete code or system architecture. This gap is further exacerbated by the politics of infrastructure, where access to computing power and advanced chips remains a privilege of certain nations, thereby marginalizing developing countries from contributing to shaping the future of technology.
Amid intense debates over governance and technical standards, an urgent need has arisen to build "middleware" or a connecting layer capable of tangibly integrating human rights principles into AI systems. Experts emphasize that high-level UN discussions will be meaningless without concrete, binding action to ensure that technology does not merely advance, but remains relevant and safe for all segments of global society.
