🔗 Share this article

Worldwide, governments are channeling enormous sums into what is known as “sovereign AI” – developing national artificial intelligence technologies. From the city-state of Singapore to the nation of Malaysia and Switzerland, countries are racing to build AI that understands local languages and local customs.

The Global AI Arms Race

This movement is part of a wider global race dominated by tech giants from the United States and China. Whereas organizations like a leading AI firm and Meta allocate enormous funds, developing countries are also making sovereign gambles in the AI landscape.

However amid such huge investments involved, is it possible for smaller states achieve notable advantages? According to a specialist from a prominent policy organization, Except if you’re a wealthy state or a major corporation, it’s a significant burden to develop an LLM from scratch.”

Defence Considerations

A lot of nations are reluctant to depend on foreign AI technologies. Across India, for example, Western-developed AI systems have occasionally fallen short. A particular case saw an AI agent employed to teach pupils in a distant village – it spoke in English with a pronounced American accent that was hard to understand for native users.

Furthermore there’s the national security factor. For India’s military authorities, using certain international AI tools is viewed not permissible. Per an developer commented, “It could have some random training dataset that might say that, for example, Ladakh is separate from India … Utilizing that particular model in a military context is a major risk.”

He continued, I’ve consulted experts who are in the military. They wish to use AI, but, setting aside certain models, they don’t even want to rely on US technologies because information might go outside the country, and that is absolutely not OK with them.”

Domestic Projects

As a result, a number of states are backing local ventures. A particular such a initiative is being developed in India, in which an organization is striving to create a sovereign LLM with public support. This effort has committed about $1.25bn to machine learning progress.

The developer foresees a system that is significantly smaller than leading tools from Western and Eastern tech companies. He notes that India will have to make up for the resource shortfall with skill. “Being in India, we lack the advantage of pouring huge sums into it,” he says. “How do we compete against for example the hundreds of billions that the United States is pumping in? I think that is where the core expertise and the strategic thinking plays a role.”

Native Priority

Throughout the city-state, a public project is supporting language models educated in the region's local dialects. These dialects – such as the Malay language, the Thai language, Lao, Indonesian, Khmer and additional ones – are frequently underrepresented in American and Asian LLMs.

I wish the people who are creating these national AI models were informed of how rapidly and how quickly the leading edge is advancing.

An executive involved in the program says that these models are intended to supplement more extensive systems, as opposed to displacing them. Tools such as a popular AI tool and another major AI system, he states, frequently struggle with local dialects and cultural aspects – speaking in stilted Khmer, as an example, or suggesting pork-based recipes to Malay users.

Developing native-tongue LLMs permits local governments to include cultural sensitivity – and at least be “informed users” of a advanced system developed elsewhere.

He adds, “I’m very careful with the term national. I think what we’re trying to say is we aim to be more adequately included and we want to comprehend the abilities” of AI platforms.

International Partnership

For nations trying to find their place in an escalating worldwide landscape, there’s another possibility: collaborate. Experts affiliated with a well-known institution put forward a state-owned AI venture allocated across a consortium of developing countries.

They call the project “an AI equivalent of Airbus”, drawing inspiration from Europe’s productive initiative to build a rival to Boeing in the mid-20th century. This idea would see the creation of a state-backed AI entity that would pool the capabilities of different states’ AI programs – such as the UK, Spain, Canada, the Federal Republic of Germany, the nation of Japan, the Republic of Singapore, South Korea, the French Republic, the Swiss Confederation and the Kingdom of Sweden – to create a strong competitor to the Western and Eastern giants.

The primary researcher of a report setting out the concept says that the concept has drawn the interest of AI leaders of at least several nations up to now, along with several state AI organizations. Although it is now centered on “middle powers”, less wealthy nations – the nation of Mongolia and the Republic of Rwanda for example – have additionally expressed interest.

He comments, Currently, I think it’s just a fact there’s reduced confidence in the assurances of the existing US administration. People are asking like, should we trust these technologies? What if they decide to