When it comes to technology and media, Singapore’s small market often means the country is a price taker – it pays what World Cup TV rights owners demand and it can’t set rules without worrying about alienating Big Tech firms. Yet, in the global AI race, it is forging ahead as a early leader.
The best examples come this week at the Asia Tech x Singapore annual technology festival, which showed the Little Red Dot’s AI progress to global industry leaders.
Among the developments, a few stand out. First, improvements to the Singapore-made Meralion multimodal large language model (LLM) aimed at Southeast Asia mean it can now process “emotional intelligence” as well as more regional languages, such as Malay, Tamil, Thai, Bahasa Indonesia and Vietnamese.
Whether used for reminding seniors to take their medication or for a company’s customer service, the LLM is said to be tuned to people in the region.
For starters, it can code-switch effectively, say, to understand Singlish, the spoken English in Singapore that includes many local dialects and languages.
Announcing the Meralion update on May 28, Singapore’s Digital Development and Information Minister, Josephine Teo, said it could even handle non-verbal cues such as a speaker’s volume, emotion and tone.
She added that the LLM could serve the more than 450 million people in the region who use various different languages to communicate every day.
Developed by Singapore’s A*Star Institute for Infocomm Research, Meralion has already attracted more than 90,000 downloads from research labs, service providers and academics since its launch in December 2024.
Now, a consortium made up of government agencies and businesses, such as DBS, Grab and Microsoft Singapore, will use it to develop practical AI applications, from multilingual customer support to health and emotional insight detection and agentic decision-making systems.
More than just finding a niche in AI for the region, Singapore is also connecting to key partners around the world to make sure it is plugged in as well as influencing the development of the most important technology breakthrough in a generation.
In AI safety, a testing framework developed here, called AI Verify, now supports generative AI and seeks to include features set out by the respected National Institute of Standards and Technology (NIST) in the United States.
This alignment with the NIST’s AI Risk Management Framework helps Singapore keep in touch with the world’s largest markets and possibly influence them with insights gleaned from the city-state’s early adoption of AI.
Of course, Singapore alone cannot enact regulations such as the European Union’s AI Act that impact countries globally or force technology companies to follow strict rules such as the trade bloc’s earlier General Data Protection Regulation (GDPR).
However, it can build cross-border links that will help businesses here to more easily adopt AI and comply with AI usage overseas. These are crucial to building up its AI hub ambitions.
This week, Singapore’s AI Safety Institute inked a deal to work with France’s AI Safety Institute to kickstart collaboration and cooperation on AI safety and cooperation.
Also this week, the new Global Cross-Border Privacy Rules (CBPR) certification announced here means businesses can more easily comply with data protection standards when they transfer data across nine economies with about 40 trillion in market size.
The certification was developed by the Global CBPR Forum, of which Singapore serves as deputy chair. And this new global certification is built on the Asia-Pacific Economic Cooperation (Apec) CBPR certification, which the likes of OCBC is certified for and can expect to benefit from.
Of course, signing agreements and connecting to the world can only do so much. Making use of AI every day is what’s ultimately crucial to a country’s AI and innovation readiness.
Some 7,200 Singapore businesses have embraced AI-enabled, integrated and cloud-based solutions in the past year, according to the Infocomm Media Development Authority (IMDA) this week.
And more professionals will be trained to use AI, with about 800 new training opportunities and up to 500 new projects to benefit 1,000 enterprises to be rolled out, reported The Straits Times.
This week, the IMDA said another new 400 training places for AI roles will be available at Amazon Web Services, Microsoft, Oracle and Singtel. The key is in developing a talent pipeline to boost AI capabilities.
To be sure, there will be limitations to Singapore’s ambitions. Learning AI may not be easy as the technology moves at such a breakneck pace – see DeepSeek early this year. Plus, the country’s size is always a constraint when it comes to expenditure on AI.
That said, by doubling down on this AI adventure, especially in these early days, Singapore is investing smartly for a future that will nobody can safely predict.
A national computerisation drive in the 1980s and 1990s ensured that the Republic was plugged into the digital world early. Now, pressing the advantage it has earned to punch above its weight in AI is the smart move. Actually, it’s the necessary move if it wants to stay relevant.
