Governance and compliance frameworks for responsible artificial intelligence deployment

The foreseeable future will likely see a profound impact caused by the advancing technology of Artificial Intelligence systems, in our day-to-day activities and routines, just like the same effect brought about by the introduction of the internet and World Wide Web [1-2]. The AIs currently in use, and the coming future of general intelligence machine systems, demand and expect appropriate – or even the best possible – Governance and Compliance Laws to be established to avert any possible negative consequences that could arise from their interaction with humanity, nature, the economy, governments, and society [2-4]. In other words, they need to answer two vital questions: who governs who? And who governs what? That said, the question of Governance indeed takes centre stage.

AI Governance refers to the organizational structures, control mechanisms, and decision-making powers that determine how AI technology is created and used, as well as its impact on global society and the economy. Unlike most other technologies, governance tends to be sub-optimal: rules are allowed to mushroom, gaps are left unaddressed, and overlapping authorities create confusion. For normative reasons, we want to prohibit the risks associated with catastrophically bad outcomes: superintelligent AIs programmed to achieve unfortunate goals, autonomous weapons likely to escalate conflicts and create battlefields free of human decision-making, and decision-support systems that contribute to the emergence of dystopian society through massive discrimination, polarization, and deception. For practical purposes, we also want to ensure that AI serves social and business goals: autonomous cars and delivery drones perform safely and efficiently; AI-enhanced medical diagnoses and digital education improve the quality of life; and intelligent transaction management systems add value to companies, investors, and customers.