ASTANA – Artificial intelligence (AI) is no longer a concept from futuristic films. It is here – reshaping economies, industries and the way societies function. While some countries are racing ahead, others are figuring out how to navigate this uncharted territory. The Astana Times spoke to Richard Stirling, head of Oxford Insights, a global research and consulting firm behind the Government AI Readiness Index, to discuss what it takes to be AI ready, the role of data, and the underlying risks.
Stirling has been at the intersection of government and technology throughout his career, advising centres of government on a broad range of areas. Photo credit: oxfordinsights.com
Stirling visited Almaty, Kazakhstan’s largest city, at the end of January to attend the Digital Almaty. This major technology conference brought together policymakers, tech leaders and investors from across Central Asia to explore AI’s impact on the governance and cybersecurity industry.
From his discussions with Kazakh officials and tech leaders, Stirling witnessed how the nation is making strategic moves to position itself as a digital frontrunner in Central Asia.
“It is clear that there is a really strong vision and ambition,” Stirling said.
“The types of things that I’ve seen are quite impressive, particularly around AI. The international AI center, which they are setting up; the training program, which they are doing in partnership with Google; and the education program that was launched at Digital Almaty. This is a pretty comprehensive suite of initiatives to build digital capacity, and particularly AI capacity in the country. Time will tell how it plays out, but it is looking pretty good at the moment,” he said.
What is an AI ready government?
When it comes to AI readiness, Stirling is clear – it is more than just having access to cutting-edge technology.
“An ideal AI ready government is one that has a strong performance in government, in the technology sector and the data and infrastructure,” Stirling said.
Oxford Insights publishes the Government AI Readiness Index annually, a global benchmark assessing how well different nations are positioned to harness AI for the public good. In its latest edition, where Kazakhstan ranked 76th, the index analyzed 188 countries across 40 indicators in three core pillars – government, technology sector, and data & infrastructure.
“In the government, you need to have a clear vision of what you want to happen with AI and how to harness it,” Stirling added. It also entails appropriate regulation and attention to ethical risks.
“You need a domestic technology sector which is able to work with AI and harness it. And the final thing is – you need the data and the digital infrastructure to go with it because without having the right data in the economy, then there is nothing for the AI to run on, so you need to place the data infrastructure to support it,” Stirling said.
The nations that score the highest are the United States, Singapore, and South Korea. While the U.S. leads the ranking, Stirling sees Singapore as the most AI ready government because of its “good balanced performance.”
The U.S.’s primary strength lies in the technology sector pillar, where it benefits from a significantly larger and more developed market. Its lead in other areas is less pronounced.
“They [Singapore] have taken steps to work out what they want to achieve with AI and how they will put guardrails around it. They have a very strong technology sector, which they’ve been investing in for the last 20-30 years, including with quite an interventionist state,” he said.
Singapore also has a solid data infrastructure. “They have worked through all of the issues around data collection, cleaning, accessibility, and making sure that everything has an API [Application Programming Interface]. It is a very digitized economy,” he said.
AI market is still early
Stirling sees the market for AI as still early.
“I think the global AI market will really come alive when the use cases become more mainstream. When we see AI being used in all of the sectors and industries, because that’s at the point when it goes from being a really interesting proof of concepts tool into something which is much more mainstream,” Stirling said.
Opportunities for Kazakhstan
With its growing tech sector and government-backed digital initiatives, Kazakhstan is exploring how AI can support its economic development.
In the education sector, the nation recently launched the AI Sana program to develop AI skills among students. With several stages, from an advanced curriculum with lectures and seminars to business acceleration, the program’s first stage aims to reach 100,000 participants.
“There is an opportunity to increase the digital skills within the economy. That’s reflected in the initiative that Minister Nurbek [Minister of Education Sayasat Nurbek] launched last week around giving AI skills to children and embedding AI in the education of the nation’s youth. But you can also see that more widely in the economy,” he said.
Stirling pointed out the need for a skilled workforce ready to harness the potential of AI fully.
“Fortunately, Kazakhstan already has good maths and computer science skills, so there is a good technical base,” he added.
Many of the necessary initiatives—such as establishing startup incubators, providing office spaces, and attracting investment—are being implemented by the government. When asked about further measures, Stirling suggested advertising the success stories.
“Often, you’re really good at doing the thing and less good at shouting about it. If what you’re looking to do is attract people from outside to come and set up in Kazakhstan, you need to shout about the successes of the people who’ve done that,” he added.
The role of data
Data is one of the most significant hurdles in AI integration, and Stirling explains why it remains a foundational pillar for building AI capacity.
“Before Christmas, the prevalent narrative was you can build an ever more powerful AI or LLM [large language model] if you have more data and more data centers. Basically, it is all about compute and the availability of data. Deep Seek showed that there is still some space left for innovation in the maths and algorithms. You still need the other two pillars,” he said
Without clean, accessible data, AI systems can’t reach their full potential, leading to flawed algorithms and, worse, biased decision-making.
“The other role for data in Kazakhstan or other countries is you should publish the data so that the AI is built in a way that reflects your society. If you took an extreme case and you published no data about Kazakhstan, none of the AIs that are built or LLMs that are built would be able to understand anything about Kazakhstan, the history, the culture, and it wouldn’t get reflected. (…) Therefore, as people start using AIs, which they will do because they are really useful, they will be getting answers that were from someone else’s perspective or were partial. You need to publish the data to fill in that gap,” he said.
