🔗 Share this article

Worldwide, nations are pouring enormous sums into the concept of “sovereign AI” – developing national artificial intelligence technologies. From the city-state of Singapore to the nation of Malaysia and Switzerland, states are vying to build AI that understands regional dialects and cultural specifics.

The International AI Competition

This movement is a component of a larger international race led by large firms from the US and China. While companies like a leading AI firm and Meta allocate enormous funds, mid-sized nations are also taking sovereign bets in the AI landscape.

However amid such vast investments involved, can smaller nations secure significant gains? According to a specialist from an influential thinktank, Except if you’re a wealthy government or a large company, it’s a substantial hardship to develop an LLM from scratch.”

Defence Concerns

Many countries are hesitant to rely on overseas AI technologies. Across India, for instance, American-made AI tools have at times fallen short. A particular instance involved an AI agent employed to instruct pupils in a isolated village – it spoke in the English language with a pronounced Western inflection that was nearly-incomprehensible for regional students.

Then there’s the defence dimension. In the Indian defence ministry, using particular external AI tools is considered unacceptable. According to a developer commented, There might be some arbitrary learning material that might say that, oh, a certain region is outside of India … Employing that particular system in a military context is a serious concern.”

He continued, I’ve discussed with people who are in defence. They wish to use AI, but, forget about specific systems, they prefer not to rely on Western systems because information may be transferred outside the country, and that is completely unacceptable with them.”

Homegrown Efforts

In response, a number of nations are funding local ventures. One this project is in progress in the Indian market, wherein a company is attempting to develop a domestic LLM with state funding. This project has committed approximately $1.25bn to AI development.

The developer envisions a system that is less resource-intensive than premier models from US and Chinese tech companies. He explains that the country will have to compensate for the resource shortfall with expertise. Based in India, we lack the luxury of pouring huge sums into it,” he says. “How do we contend against say the $100 or $300 or $500bn that the America is devoting? I think that is the point at which the fundamental knowledge and the strategic thinking plays a role.”

Local Emphasis

Throughout the city-state, a public project is funding AI systems developed in local regional languages. Such dialects – such as the Malay language, the Thai language, the Lao language, Bahasa Indonesia, the Khmer language and additional ones – are often underrepresented in Western-developed LLMs.

I hope the individuals who are building these sovereign AI systems were conscious of how rapidly and how quickly the leading edge is progressing.

A senior director participating in the project notes that these tools are created to complement more extensive models, instead of displacing them. Tools such as a popular AI tool and Gemini, he states, commonly have difficulty with local dialects and local customs – interacting in unnatural Khmer, for example, or recommending non-vegetarian recipes to Malay users.

Building native-tongue LLMs enables national authorities to code in local context – and at least be “smart consumers” of a powerful technology built overseas.

He adds, I am cautious with the word national. I think what we’re attempting to express is we wish to be more adequately included and we aim to comprehend the capabilities” of AI systems.

Cross-Border Collaboration

Regarding nations attempting to establish a position in an escalating global market, there’s a different approach: join forces. Researchers connected to a respected policy school recently proposed a public AI company distributed among a alliance of developing nations.

They call the initiative “an AI equivalent of Airbus”, modeled after the European successful strategy to build a rival to Boeing in the 1960s. Their proposal would entail the establishment of a government-supported AI organization that would pool the assets of several nations’ AI projects – including the United Kingdom, Spain, Canada, the Federal Republic of Germany, Japan, Singapore, South Korea, the French Republic, Switzerland and Sweden – to establish a strong competitor to the US and Chinese major players.

The lead author of a study describing the proposal notes that the idea has attracted the consideration of AI leaders of at least three states to date, in addition to multiple state AI firms. Although it is presently targeting “middle powers”, developing countries – the nation of Mongolia and Rwanda included – have additionally indicated willingness.

He comments, In today’s climate, I think it’s an accepted truth there’s diminished faith in the promises of the existing White House. Individuals are wondering such as, is it safe to rely on any of this tech? Suppose they decide to