A warm welcome to esteemed co-hosts, Mr Cho and Mr Kim, as well as the Defence Ministers and Excellencies. Let me thank the Republic of Korea for your hospitality and acknowledge your leadership in convening this event. We are very honoured to co-host the 2^nd REAIM Summit, together with you, the Netherlands, Kenya, and the United Kingdom.

Benefits and Risks of AI

Mr Cho set us on reminiscing of missed opportunities and hindsight. I have another aphorism for you – “If I could do it all over again, I may have done it differently”. This is a familiar phrase that many of us through our own life’s journey have uttered from time to time. I think as it does for individuals, this phrase also applies to organisations, societies and even our global world.

Here’s a thought experiment. What would we have done differently then, with the benefit of hindsight in the global wave of digitalisation over the last two decades, if we could do it all over again? Many of us would agree on obvious aspects that stand out. Certainly, we would have put in more resources to narrow the digital divide, even requiring corporations as part of their submission for tender contracts to provide these digital tools to the poor and underserved. For social media platforms, we certainly would have erected restrictions and safeguards to require app producers to have built-in protection against misinformation and other safeguards.

But of course, we found out how difficult it is to try to tackle these present dilemmas in the digital domain, when business models have been built on mass consumption with inadequate safeguards. Some would argue quite convincingly that the horse has bolted for this domain.

Which brings us to the reason why this 2^nd REAIM Summit has been convened. I want to thank the Republic of Korea and the Netherlands for beginning this much needed and bold endeavour. All of us know that AI, just like digitalisation, will change and even disrupt our lives. AI adoption and growth is exponential. From now till 2030, the global GPU[1] market as a proxy, is estimated to grow by a Compound Annual Growth Rate of more than 30% to more than US$400B.

But if there is any lesson to be learnt from the ill-effects of unfettered development in the digital domain, it must be this – it will be that much more effective to have consensus and regulation early in the development of any new technology, and this particularly applies to the adoption of AI in the military sphere. Because if things go badly in the AI-enabled military platforms, whether conventional or nuclear, the consequences can be catastrophic, devastating and even existential.

Importance of the REAIM Process

The REAIM process has provided this timely and needed platform to decide on a consensus for governance of AI in the military domain. The goal is clear and laudable but the obstacles are aplenty. Particularly because all countries will want to maximise the application of AI to gain an edge over their adversaries. An AI governance framework is required to protect our humanity and global world.

Singapore congratulates the inaugural REAIM Summit last year that attracted 57 countries, including the US and China, to endorse the REAIM ‘Call to Action’. In February this year, my Ministry felt compelled to support this good start and to keep the momentum going by hosting the regional consultations for Asia, with the Republic of Korea and the Netherlands. We were glad to have representatives from 15 countries, including our fellow ASEAN members, China, and India, to discuss these challenges.

My co-hosts and I hope that this year’s REAIM Summit has been useful in allowing us all to build a consensus on the REAIM ‘Blueprint for Action’. This Blueprint represents a step forward by providing us with action-oriented guidelines to implement responsible AI within our militaries, ensuring appropriate human decision making at crisis points.

We hope that by following this REAIM process, rather than “wish I had done it differently”, we will all be able to say, “I’m glad that we started right and pre-empted problems.” Thank you.