Should Artificial Intelligence Governance be Centralised?

Abstract

Can effective international governance for artificial intelligence remain fragmented, or is there a need for a centralised international organisation for AI? We draw on the history of other international regimes to identify advantages and disadvantages in centralising AI governance. Some considerations, such as efficiency and political power, speak in favour of centralisation. Conversely, the risk of creating a slow and brittle institution speaks against it, as does the difficulty in securing participation while creating stringent rules. Other considerations depend on the specific design of a centralised institution. A well-designed body may be able to deter forum shopping and ensure policy coordination. However, forum shopping can be beneficial and a fragmented landscape of institutions can be self-organising. Centralisation entails trade-offs and the details matter. We conclude with two core recommendations. First, the outcome will depend on the exact design of a central institution. A well-designed centralised regime covering a set of coherent issues could be beneficial. But locking-in an inadequate structure may pose a fate worse than fragmentation. Second, for now fragmentation will likely persist. This should be closely monitored to see if it is self-organising or simply inadequate.

Policy brief: the creation of a G20 coordinating committee for the governance of artificial intelligence

“…It has also been accompanied by responses from international and multistakeholder organizations, as well as civil society. Those rather distributed or decentralized, international AI governance structures can be divided into governance networks and polycentric governance [5,36]. While the existing cyber regime complex is centered around international organizations and different regime complexes (cf.…”

“…Therefore, governments and other stakeholders face a dysfunctional international cyber regime complex [27,23]. The more recent demand for digital sovereignty raised by various governments that seek a balance between protection and collaboration risks leading to digital nationalism, further fragmenting the cyber regime complex and rendering emerging local and regional AI governance approaches ineffective [5,26]. Thus, a comprehensive, coordinated global approach seems to be effective in countering such fragmentation and a lack of collaboration, and it is instrumental to planning and preparing against, mitigating, and recovering from future threats and structural imbalances [42].…”

“…In addition, it is criticized as lacking meaningful representation from small nations and underserved communities that are most vulnerable to AI-driven disruptions. The group has the capacity for such global stewardship, but such a mandate would demand further institutionalization of the G20 process, including the reform and introduction of new functional and organizational features [5].…”

“…Such a role would build upon and align with established procedures, shared longterm orientations and action plans, and joint presentations and appearances. Implementing a CCGAI would require, at a minimum, the following four functional features, which would strengthen the committee's mandate as a "metagovernor" [20]: coordination, accountability, foresight, and consultation [1,5,15,34,35].…”

“…Foresight information should be stored in the existing G20 Repository of Digital Policies [13]. d. Public consultation and inclusivity improve the transparency and effectiveness of the governance coordination process and create legitimacy and trust [1,4,5]. A consultation mechanism must be formalized in which stakeholders, especially civil society groups and non-G20 countries, are integrated into a separate secretariat and contribute at the official policy discussion level.…”