🔗 Share this article

Worldwide, nations are investing hundreds of billions into what's termed “sovereign AI” – developing national AI technologies. From the city-state of Singapore to Malaysia and Switzerland, nations are vying to develop AI that grasps local languages and cultural specifics.

The International AI Competition

This trend is part of a larger global race led by tech giants from the US and the People's Republic of China. Whereas organizations like a leading AI firm and a social media giant invest massive funds, mid-sized nations are also taking sovereign gambles in the AI field.

However given such huge investments at stake, can developing countries secure meaningful advantages? As stated by a analyst from a prominent research institute, “Unless you’re a affluent nation or a major corporation, it’s quite a hardship to develop an LLM from scratch.”

Defence Concerns

Many countries are hesitant to use foreign AI models. In India, as an example, US-built AI solutions have occasionally been insufficient. One case saw an AI tool deployed to instruct students in a distant village – it communicated in English with a pronounced American accent that was nearly-incomprehensible for native users.

Then there’s the national security aspect. For India’s defence ministry, relying on certain foreign systems is viewed inadmissible. According to a entrepreneur explained, “It could have some random learning material that may state that, for example, a certain region is separate from India … Employing that certain AI in a security environment is a serious concern.”

He further stated, “I have spoken to experts who are in security. They wish to use AI, but, disregarding certain models, they are reluctant to rely on American systems because information may be transferred outside the country, and that is totally inappropriate with them.”

Domestic Projects

As a result, a number of states are supporting national projects. An example such a initiative is being developed in the Indian market, where a firm is attempting to create a sovereign LLM with government backing. This project has dedicated roughly 1.25 billion dollars to machine learning progress.

The founder foresees a model that is significantly smaller than premier systems from Western and Eastern firms. He states that the country will have to make up for the resource shortfall with skill. Located in India, we lack the option of pouring massive funds into it,” he says. “How do we compete against for example the $100 or $300 or $500bn that the US is pumping in? I think that is where the fundamental knowledge and the brain game plays a role.”

Native Emphasis

Across Singapore, a government initiative is backing language models developed in south-east Asia’s regional languages. These dialects – including the Malay language, the Thai language, Lao, Bahasa Indonesia, the Khmer language and others – are commonly underrepresented in American and Asian LLMs.

I hope the people who are developing these sovereign AI systems were informed of just how far and how quickly the cutting edge is progressing.

A senior director involved in the initiative explains that these systems are designed to supplement bigger models, instead of replacing them. Tools such as a popular AI tool and another major AI system, he says, commonly struggle with local dialects and local customs – communicating in awkward Khmer, as an example, or suggesting meat-containing recipes to Malaysian users.

Creating native-tongue LLMs enables state agencies to incorporate cultural sensitivity – and at least be “knowledgeable adopters” of a advanced tool developed elsewhere.

He further explains, I am cautious with the term independent. I think what we’re trying to say is we want to be better represented and we want to understand the abilities” of AI platforms.

Cross-Border Partnership

For countries seeking to establish a position in an growing global market, there’s an alternative: team up. Analysts associated with a well-known policy school recently proposed a state-owned AI venture shared among a group of middle-income states.

They refer to the proposal “a collaborative AI effort”, drawing inspiration from the European productive play to build a competitor to Boeing in the mid-20th century. Their proposal would involve the creation of a government-supported AI organization that would pool the assets of different states’ AI initiatives – for example the UK, the Kingdom of Spain, the Canadian government, the Federal Republic of Germany, the nation of Japan, Singapore, South Korea, France, the Swiss Confederation and the Kingdom of Sweden – to create a viable alternative to the US and Chinese leaders.

The primary researcher of a report setting out the concept says that the proposal has gained the consideration of AI ministers of at least several countries so far, in addition to several national AI organizations. Although it is now targeting “mid-sized nations”, less wealthy nations – Mongolia and Rwanda among them – have also shown curiosity.

He comments, “Nowadays, I think it’s an accepted truth there’s diminished faith in the promises of the present US administration. Experts are questioning for example, should we trust such systems? In case they decide to