Singapore Leads Sovereign AI Race
Singapore Leads Sovereign AI Race
Synopsis:
- Governments worldwide are investing billions to build their own “sovereign AI” models – Guardian report.
- Singapore’s SEA-LION, India’s Soket AI, and Malaysia’s ILMUchat lead regional development efforts.
- Experts warn smaller nations may struggle to compete with US and Chinese AI giants.
- Some analysts suggest focusing on AI regulation and cooperation instead of replication.
Estimated Reading Time: 3 mins
Governments Push for National AI Independence
In a report by the Guardian, around the world, governments are channeling billions into developing their own artificial intelligence models under the growing concept of “sovereign AI.” The strategy aims to ensure national control over data, language, and cultural nuances while reducing reliance on systems built by US and Chinese tech giants such as OpenAI, Meta, and Alibaba.
In Singapore, a government-funded AI program supports the SEA-LION family of models, capable of conversing in 11 Southeast Asian languages—from Bahasa Indonesia to Lao. Malaysia’s ILMUchat, developed by a local construction conglomerate, can differentiate between Georgetown, Penang, and Georgetown University in the US. Meanwhile, Switzerland’s Apertus, unveiled in September, distinguishes when to use the Swiss German “ss” instead of the German “ß.”
These initiatives are part of a broader movement as middle-income and smaller nations attempt to stake their claim in an AI landscape worth hundreds of billions of dollars.
Smaller Nations Struggle Against Superpower Giants
“While the US, China, and their leading companies can blitzkrieg their way into AI dominance, middle powers face a harder battle,” says Trisha Ray, resident fellow at the Atlantic Council. “Unless you’re a wealthy government or major corporation, building a large language model from scratch is a heavy burden.”
India, the second-largest market for OpenAI globally, illustrates the dilemma. Despite over 100 million ChatGPT downloads, many local developers say US-made systems fail to capture India’s linguistic and contextual realities. Soket AI founder Abhishek Upperwal cites an AI tutor in Telangana speaking with an almost incomprehensible American accent and another legal startup’s failed attempt to adapt Meta’s LLaMa, which produced a confusing mix of US and Indian legal frameworks.
Security concerns add another layer. “Using Chinese models like DeepSeek is off the table,” Upperwal explains. “It could include data suggesting Ladakh isn’t part of India—completely unacceptable in defence applications.”
Supported by India’s $1.25 billion IndiaAI Mission, Soket AI aims to build a national-scale LLM smaller than those of global leaders like Mistral but optimized for local use. “We don’t have billions to spend,” says Upperwal. “Our advantage must come from talent and ingenuity.”
Singapore’s SEA-LION and the Regional Language Gap
Singapore’s government initiative AI Singapore backs SEA-LION, a series of LLMs trained on Southeast Asia’s underrepresented languages, including Malay, Thai, Khmer, and Bahasa Indonesia.
Leslie Teo, senior director at AI Singapore, says these models are designed to complement, not replace, systems like ChatGPT or Gemini. “Global models often mishandle regional nuances—like formal Khmer or recommending pork recipes to Muslim users,” he notes. “We want to ensure our cultural and linguistic diversity is represented, and that we’re smart consumers of these global technologies.”
Teo stresses that “sovereign” AI doesn’t necessarily mean isolationism but awareness. “We want to understand AI’s capabilities and limitations so we can engage on equal footing.”
Cooperation: An “Airbus for AI” Proposal
Beyond national efforts, some researchers are calling for multinational collaboration. Scholars at Cambridge’s Bennett School for Public Policy have proposed an “Airbus for AI”, inspired by Europe’s successful aerospace consortium that rivaled Boeing.
The plan envisions a joint public AI company pooling resources from countries like the UK, Spain, Canada, Japan, Singapore, South Korea, France, Switzerland, and Sweden to compete with US and Chinese powerhouses.
Joshua Tan, the paper’s lead author, says the idea has attracted attention from ministers in at least three nations and several sovereign AI companies. Even smaller economies, including Mongolia and Rwanda, have shown interest. “There’s growing distrust in relying solely on US-based technologies,” Tan says. “What happens if they decide to switch off access?”
Experts Warn of Costly Missteps
Despite the enthusiasm, some experts remain skeptical. Malaysian AI strategist Tzu Kit Chan, who advises his government, warns that the pace of AI development makes replication risky and expensive. “Many underestimate how far and fast the frontier is moving,” he cautions. “Governments risk wasting vast sums if they pursue outdated strategies.”
Instead of competing head-on with industry giants, Chan argues nations should invest in AI safety and regulatory frameworks to ensure responsible adoption. “Walk the streets of Kuala Lumpur and ask any finance professional what they use. Eight out of ten will say ChatGPT or Gemini—not a sovereign model.”
Conclusion
The rise of sovereign AI reflects both ambition and anxiety: a desire for technological independence coupled with the reality of global inequality in compute and capital. As Singapore, India, and Malaysia push forward, the question remains—can sovereign AI truly deliver national advantage, or will it become another costly experiment in chasing a moving frontier?
Source here – Have a Story? Address it to the Editor and submit it here
About AI Singapore
AI Singapore is a national artificial intelligence initiative launched in 2017 by the National Research Foundation (NRF) under the Prime Minister’s Office of Singapore. It brings together government agencies, research institutions, and industry partners to develop AI capabilities, drive innovation, and build local expertise. The program’s mission is to anchor deep national capabilities in AI, enhance industry adoption, and nurture homegrown talent.
AI Singapore runs several flagship programs, including 100 Experiments (100E), which helps companies implement AI solutions, and AI Apprenticeship Programme (AIAP), which trains professionals for AI careers. The organization is led by Professor Ho Teck Hua as Executive Chairman and Leslie Teo as Senior Director for AI products and innovation. Through projects like SEA-LION, AI Singapore focuses on creating language models that represent Southeast Asian cultures and languages, reinforcing Singapore’s role as a regional leader in ethical, inclusive, and applied AI development.
Featured Image: Expedia
Disclaimer
The information provided in this article is for general informational purposes only and from publicly available sources. While we strive for accuracy, we do not make any representations or warranties, express or implied, regarding the completeness, reliability, or validity of the content. This article does not make any direct claims about specific companies, individuals, or organizations. Any references to reports or external sources are for context and do not imply endorsement or verification of any specific allegations. Readers are encouraged to conduct their own research and seek professional advice before making business decisions. We disclaim any liability for any losses or damages incurred as a result of reliance on the information provided.