Singapore’s Approach

During the Seoul AI Summit in May this year, Singapore’s Prime Minister Lawrence Wong highlighted our country’s National AI Strategy 2.0 to harness AI’s benefits while mitigating potential harm. Under this national blueprint, various sectors must develop detailed frameworks for AI governance to achieve those outcomes. Accordingly, for the defence sector, since 2021, we have established principles for responsible, robust, reliable, and safe AI under four broad areas:

• Accountability for outcomes of AI systems;
• Appropriate levels of testing and assurance;
• Ab initio design of the AI systems to deal with cyber and adversarial threats; and lastly
• The risks of AI failure.

International Cooperation

Singapore is not alone in developing these blueprints for safe AI. But as all of us are aware, without a global consensus, it will be hard for such countries to achieve these goals, no matter how well meaning they may be.

Thus, REAIM serves as a much-needed platform for all stakeholders, and the process involves active consultation. We are proud to co-host the regional consultations for Asia in Singapore. They complemented other regional consultations held in Latin America, Middle East, and Africa, co-hosted by our colleagues from Chile, Türkiye, and Kenya.

Arising from the consultation, REAIM is now in a position to turn the ‘Call to Action’ into detailed recommendations for this year’s ‘Blueprint for Action’. It sets up practical guidelines to implement AI responsibly through establishing safeguards, maintaining responsibility for the effects of AI systems, and ensuring that systems are developed in accordance with international law.

Singapore endorses the ‘Blueprint for Action’, as it is aligned to our national and defence sector blueprint. We encourage all here to endorse the Blueprint. Only if we work together, can there be assurance that AI will improve and not harm our economies and people.

Thank you.