As various sectors scramble to develop and adopt artificial intelligence, governments have similarly hastened the process to regulate AI to minimise perceived harms. However, AI governance has become a challenge, as it needs to balance concerns from various stakeholders and confront difficult policy questions.
This raises the question of whether domestic or international-level processes are more effective in AI governance. The Centre of Multilateralism Studies (CMS) at RSIS explored this question in a webinar held on 10 October 2024. It featured Dr Robert Trager of the Oxford Martin AI Governance Initiative, and Dr Aaron Maniam of the Oxford Blavatnik School of Government as guest speakers; Mr Benjamin Ang, Head of the Centre of Excellence for National Security at RSIS as discussant; and Dr Joel Ng, Deputy Head of CMS as chair.
Dr Trager shared insights on how best to govern civilian AI. He believes three pillars are crucial: developing forward-looking international standards, providing incentives for states to adopt those standards through specialised organisations (akin to the ICAO in aviation), and ensuring cooperation in enforcing the regime through a reporting mechanism.
Dr Maniam discussed how organisational structures at the multilateral level can determine the governance framework that will emerge from each international organisation. For example, in the UN, a culture of seeking the common ground and understanding, produces a governance framework focused on inclusivity and preserving the public good. In any case, bold thinking and leadership are needed to field innovative governance solutions.
Mr Ang focused on the emerging cybersecurity risks that AI may bring, such as in creating weapons and disrupting the financial sector, and that the public and private sector must play complementary roles to manage these risks. However, geopolitical tensions and the outsized impact of the tech sector can complicate the governance process.
After the presentations, the audience asked questions regarding the effectiveness of soft law and hard law in the context of AI regulation and whether states will find international regulatory oversight appealing if many seek the entry of AI firms into their markets.
