Government officials will never ask you to transfer money or disclose bank log-in details over a phone call.

Call the 24/7 ScamShield Helpline at 1799 if you are unsure if something is a scam.

GOH Speech by Deputy Prime Minister and Minister for Trade and Industry Gan Kim Yong at the NCS Impact Forum 2025

GOH Speech by Deputy Prime Minister and Minister for Trade and Industry Gan Kim Yong at the NCS Impact Forum 2025

Mr. Lim Swee Say, Chairman, NCS

 

Mr. Yuen Kuan Moon, Group Chief Executive Officer, Singtel

 

Mr. Ng Kuo Pin, Chief Executive Officer, NCS

 

Distinguished guests,

 

Ladies and gentlemen, good morning to all of you.

 

1. Thank you for inviting me to speak at NCS Impact 2025.

 

2. This year’s theme – ‘New Horizon: The Era of AI’ – aptly reflects where we are in the evolution of AI.

 

3. Three years after ChatGPT took the world by storm, we have seen significant advances in the capabilities, accessibility and adoption of Artificial Intelligence.

 

a. From coding assistants, to writing tools, to customer service bots, and even creative design platforms, generative AI is now a quintessential element of our digital landscape.

 

b. Multi-modal AI models that can process text, images, audio and video are reshaping how we interact with technology in real time.

 

c. AI is also increasingly being embedded into critical workflows across manufacturing, healthcare, finance and education.

 

4. These developments are driven not only by rapid advances in model capabilities, but also a significant reduction in operating costs.

 

  a. According to the 2025 AI Index Report by Stanford University, the inference costs for a system performing at the level of GPT-3.5 has dropped over 280-fold; in fact, GPT-4 class workloads are already 240 times cheaper than they were just a year ago.

 

  b. At the hardware level, costs have declined by 30% annually, while energy efficiency has improved by 40% per year.

 

  c. Open-weight models are also closing the performance gap with closed models, and now differ by only about 2% on some measures, compared to about 8% from a year ago. 

 

  d. With better hardware, more efficient architectures and greater open-source collaboration, AI has become more efficient, affordable and accessible.

 

5. We should harness the power of AI to uplift and empower our people and businesses.

 

  a. Two years ago, we launched our second National AI Strategy. We believe that by building the right capabilities within our companies and workforce, we can unlock a new frontier of economic growth through the use of AI.

 

b. We have made good progress in catalysing AI innovation within our economy.

 

  i. To date, more than 25 companies – including American Express, Grab, Oracle and Prudential – have set up AI Centres of Excellence to drive the creation of AI solutions.

 

  ii. We have also set up an AI Centre of Excellence for the Manufacturing Sector, to support the development of AI-driven solutions for common use-cases within the sector, such as predictive maintenance and product or component design.

 

  c. As part of our Smart Nation 2.0 vision, Singapore has also developed AI products across the Public Service, in areas such as homeland security, healthcare, education and environmental sustainability.

 

6. Even as we push ahead with these innovations, we must ensure that the benefits of AI are shared broadly. This will require us to institutionalise the use of AI within companies, localise AI-driven models and solutions, and professionalise our workforce with AI skills. Let me speak on each of these in turn.

 

Institutionalising AI

 

7. First, we must institutionalise the use of AI within companies big or small, by embedding it within the organisation.

 

8. Today, larger organisations tend to use AI more than smaller ones.

 

  a. Based on last year’s Singapore Digital Economy Report, AI adoption has more than doubled among large enterprises over the past five years, from about 17% in 2018 to 44% in 2023.

 

  b. On the other hand, AI uptake among SMEs increased from 3.5% to 4.2% over the same period.

 

  c. Despite the lower rate of adoption among SMEs, it is heartening to note that both SMEs and large enterprises alike saw benefits from using AI – in fact, more than 90% of both types of companies that used AI reported that the technology had improved productivity and processes.

 

9. We must double down on our efforts to help companies, especially smaller firms, benefit from AI.

 

  a. IMDA’s Generative AI Playbook for Enterprises lays out the various benefits of GenAI tools, and provides guidance on whether businesses should buy or build GenAI solutions.

 

  b. IMDA has also introduced the GenAI Navigator tool under its CTO-as-a-Service programme, which helps SMEs identify GenAI solutions that match their specific business needs.

 

10. At the same time, industry players like NCS can also help accelerate AI adoption.

 

  a. Besides offering AI-enabled solutions, they can partner companies to co-develop use cases, re-design their workflows, and even transform their business models.

 

  b. I am glad to note that NCS will invest S$130 million over the next three years to support its clients in scaling AI adoption, through its proprietary suite of tools, methodologies and accelerators. NCS will also deepen its partnerships with leading technology providers such as AWS, Google Cloud and Nvidia to raise its capabilities in AI.

 

  c. Through system integrators such as NCS, companies can deploy AI more efficiently and effectively, and in time to come, develop the capabilities to scale their own AI solutions.

 

Localising AI

 

11. Just as we need to customise AI applications for specific business contexts, we also need to localise AI models and solutions for specific linguistic and cultural contexts.

 

12. This was why AI Singapore had set out to develop a family of open-source Large Language Models called SEA-LION (or Southeast Asian Languages in One Network).

 

  a. Due to the nature of their training data, existing LLMs tend to reflect the cultural background, political beliefs and social attitudes of the Western societies.

 

  b. On the other hand, SEA-LION is trained on more content produced in Southeast Asian languages such as Thai, Vietnamese or Bahasa Indonesia. It is therefore better able to understand language and cultural nuances in the region.

 

  c. AI Singapore has recently added reasoning capabilities to the latest SEA-LION model, and plans to add other competences including voice recognition and visual recognition soon.

 

13. Localised AI models such as SEA-LION can enhance the adoption of AI in our region.

 

  a. For example, GoTo Group – which owns companies such as ride-hailing firm Gojek and e-commerce firm Tokopedia in Indonesia – is using SEA-LION to build its own AI system. Instead of training its own model from scratch, SEA-LION serves as a good foundation for GoTo to build on, thereby saving the company both training time and costs.

 

  b. NCS is also among the first companies to adopt SEA-LION. I am glad to note that NCS will be expanding the use of SEA-LION in areas such as translation of legal and compliance documents, customer engagement and to detect changes in cross-border regulations.

 

  c. I hope more businesses can leverage AI for their operations in the region, through the use of locally relevant models and applications. This will not only help companies tap into new markets in our region, but also drive the growth of the digital economy in Asia.

 

Cultivating an AI-Ready Workforce

 

14. Last but not least, we must prepare our workforce to support the AI ambitions of our companies, as well as to ensure that our workers acquire the skills to benefit from AI.

 

15. To create and implement AI, companies will need practitioners such as data scientists and AI and machine learning engineers.

 

  a. At the national level, we want to raise the number of AI practitioners to 15,000 over the next three to five years.

 

  b. In addition to our existing company-led training programmes under the TechSkills Accelerator initiative, IMDA and AI Singapore have recently launched the AI Apprenticeship Industry Programme. Through this programme, we aim to train 300 practitioners over two years in AI, data and machine learning skills, including through real-world projects.

 

16. At the same time, we must also equip the broader workforce to be AI-ready – and in this regard, companies play a key role in upskilling their workforce as they transform their businesses with AI, so that their workers will be ready to take on new roles within their organisations.

 

  a. I am glad to note that NCS has been investing in developing the AI capabilities of its workforce, and has now trained more than 10,000 of its employees globally to be AI-enabled.

 

  b. This will allow NCS to better support its clients, and ensure that its workers are themselves prepared for the AI revolution.

 

Conclusion

 

17. As we confront a more disruptive global economy driven by greater volatility and protectionism, this Forum is a timely reminder that there are still good opportunities awaiting our businesses.

 

18. By working together among Government and industry, as well as with our unions, we can institutionalise the use of AI within companies, develop AI-driven models and solutions that are locally relevant, and equip our workforce with AI skills, so that we can transform our enterprises and workers to fully harness the power of AI.

 

19. I wish you a successful Impact Forum, and I hope that today’s Forum will lead to fruitful partnerships to catalyse the next bound of AI-driven growth and opportunities for our businesses and our workers.

 

20. Thank you.

 

HOME ABOUT US TRADE INDUSTRIES PARTNERSHIPS NEWSROOM RESOURCES CAREERS
Contact Us Feedback