There is no doubt that technology will change human futures; some will be more challenging than others; the newest-on-the-bloc is Artificial Intelligence (AI). As China and the US seek to dominate this new-age interest, Muhammad Faizan Fakhar argues how countries in South Asia can hope to benefit from this burgeoning industry of opportunities for its peoples and economy.    

The 21st century has been dubbed the ‘age of artificial intelligence’, and for good reason. The development and deployment of Artificial Intelligence (AI) technologies has the potential to transform almost every aspect of human life — from healthcare, defence, communication and transportation to finance and entertainment. However, as AI becomes increasingly important, another competition is emerging between the United States and China, known as the ‘AI Race.’

The AI Race is to the 21st century what the Nuclear Race was to the 20th century — a competition for technological supremacy that will impact global socio-economic and geo-political dynamics for years to come. With Biden’s CHIPS (Creating Helpful Incentives to Produce Semiconductors) and Sciences Act (2022) aimed at curbing China’s access to advanced semiconductors, and China’s commitment to become world leader in AI by 2030, this competition for technological leadership is gaining momentum with each passing day. Superiority in AI will not only generate domination and influence, but more importantly, also give a clear lead in the military domain. As this competition between global behemoths plays out, other (developing) countries are finding it challenging to navigate this great tech war whilst protecting their own national interests; South Asian countries are no exception. Before discussing the complex ramifications of the US-China AI competition for South Asia, it is important to understand the contours and components of the evolving AI systems, and the primacy of the US and China in particular domains.

Advancement and superiority in AI systems is mainly dependent upon sophistication and perfection in two critical areas — hardware and big data resources, which play a crucial role in different aspects of AI technology.  Both China and the US have their own strengths and weaknesses in these two domains.

Significance of Chips’ Supply Chain

Hardware required for any AI system is fundamentally dependent on the core component of electronic systems — semiconductor chips. Almost every aspect of modern life revolves around semiconductor chips as they are used in a wide range of electronic devices, including smartphones, computers and automobiles. Semiconductor chips also play a vital role in the development and implementation of AI systems in terms of memory capacity, computational power, and parallel processing.

However, semiconductor chips used for AI systems are highly advanced and socialised with a grading of below 10nm. The US  holds a clear lead over China in the domain of these semiconductor chips as the majority of the semiconductor supply chain is currently dominated either by the US or its allies including Japan, Taiwan, South Korea and Netherlands. This gives the US a significant advantage in the development and production of AI hardware.

Semiconductor supply chain consists of several components, including raw material, chip design and manufacturing/FABs (fabrication facilities), testing, packaging and distribution.  A break-down of each of these components shows that the US currently dominates the global supply chain of the semiconductor chips as listed below:

• Raw Materials: Raw materials for semiconductor chips include silicon wafers, chemicals and gases. The industry is led by companies such as Shin-Etsu Chemical  and Sumco Corporation (both in Japan) and Air Liquide (France, but also has a significant presence in the US) are major suppliers of these materials.
• Chip Design: The US is ahead of everyone else in chip design, with companies such as Intel, AMD and Qualcomm (all in the US) leading the way.
• Fabrication: AI systems require the most advanced (below 10 nanometres) semiconductors which are manufactured through a process known as Extreme Ultraviolet (EUV) Lithography. Netherland’s ASML is currently the market leader in lithography equipment, with a market share of around 85 per cent. On the other hand, Taiwan has the biggest manufacturing capacity of nano-chips with a market share of 92 per cent while South Korea has the remaining 8 per cent. Taiwan Semiconductor Manufacturing Company (TSMC) is the world’s largest semiconductor foundry.
• Testing and Packaging: The testing and packaging of semiconductor chips is also led by companies based in the US and its allies. Advantest (Japan) and Teradyne (US) are major players in the chip-testing industry, while ASE Technology (Taiwan) and Amkor Technology (US) are the big ones in the chip-packaging industry.
• Distribution: The distribution of semiconductor chips is a global industry, with companies based in the US, Europe and Asia. Arrow Electronics and Avnet (both in the US) are leaders in the distribution industry.

Semiconductor chips (essential for AI systems) rely upon three things: EUV lithography equipment, chip designing, and fabrication units. All three are being closely monitored and guarded by the US at the moment (as also evident by the US CHIPS and Science Act (2022) mentioned above), which is a new US export control measure aimed at limiting China’s access to semiconductors. The Act imposes restrictions on foreign semiconductor manufacturers from selling certain chips, designs, software and EUV lithography equipment to China, and simultaneously encourages foreign manufacturers to invest in US facilities by providing up to US$280 billion in finances to companies.

Access to Data Resources and AI

While semiconductor chips form the core component of AI systems’ hardware, big data serves as the key to AI training and optimisation. Access to data resources is, therefore, equally important in creating superior AI systems. China has a population of 1.4 billion people, generating a massive amount of data every day, giving Chinese companies and government agencies access to a large pool of data. The Chinese government has made significant investments in developing data infrastructure, including building data centres and promoting data-sharing, thus creating a favourable environment for AI development; companies like Alibaba, Baidu and Tencent have access to large amounts of data from their vast consumer base, giving them an advantage in developing AI algorithms and applications. China has also made progress in developing AI-ready data for specific industries like healthcare, finance and transportation, which can provide insights into complex real-world problems.

Similarly, as part of the Digital Silk Road initiative, China is investing in the development of digital infrastructure such as fibre-optic cables, data centres and other digital platforms in countries along the Belt and Road Initiative (BRI). This infrastructure is designed to facilitate digital trade and investment. China’s investments in digital infrastructure along the BRI can also enable Chinese companies to gain greater access to data from around the world, which could provide insights into local consumer behaviour, market trends and other valuable information. Access to such unique data resources can ultimately give an advantage to China in training and optimisation of AI systems.

At the same time, the US also has access to high quality data with many of the world’s leading AI companies (like Google, Amazon and Microsoft, all based in the US). The US also has a significant amount of publicly available data, such as government data and academic research, which can be used to develop AI algorithms and applications.

In this game, China might have an advantage in terms of access to a larger volume of data.

Potential Risks, Challenges and Benefits for South Asia

This AI race between global giants poses significant socio-economic and security challenges for developing nations. Countries that fail to keep up with emerging AI-driven trends are likely to face setbacks that could hinder their progress.  There is, however, a silver lining in this competition as it offers numerous lucrative opportunities to countries in South Asia, particularly India and Pakistan.

South Asia is home to a quarter of the world’s population, with a combined population of nearly 1.9 billion people. India has become a critical strategic ally of the US in the region, while Pakistan also has a long-standing history of diplomatic, strategic and economic relations with both China and the US. Emerging strategic alliances in the Indo-Pacific (such as QUAD and AUKUS) have brought the US-China competition to the forefront of South Asian politics.

India and Pakistan have the potential to gain multiple benefits by positioning themselves as part of the global supply chain of microchips and data resources. As mentioned above, China and the US are actively pursuing tech superiority, creating a high demand for advanced technologies to satisfy the needs for AI systems and big data analytics. By developing capabilities in these areas, South Asian countries can become valuable contributors to the global tech ecosystem, creating jobs and improving the quality of life for their citizens.

India’s Ingress in the Semiconductor Supply Chain

India has already launched a number of policies and initiatives aimed at tapping the supply chain of semiconductor chips.  In November 2020, the Indian government announced the Production Linked Incentive (PLI) Scheme under which it was announced that India would invest US$963 million to incentivise semiconductor manufacturing in the country. Similarly, in January 2021, it launched the Semiconductor Fabless Accelerator Lab (SFAL) initiative, with the aim of promoting domestic design and development of semiconductor chips.

Overall, India’s strong domestic demand, skilled workforce, favourable government policies, strategic location, and low labour costs are several factors that can help the country become a significant player in the global semiconductor supply chain. Foreign investments are also playing a key role in India’s semiconductor manufacturing ecosystem with companies such as Intel, Qualcomm, Samsung and TSMC pouring billions of dollars into India.

However, India will have to address several key challenges related to infrastructure, skilled manufacturing manpower, costs, and research and development in order to realise the goal of becoming a self-sufficient semiconductor manufacturer.

Potential of Pakistan

Pakistan has not yet entered the AI market in terms of semiconductor supply chain and data resources. Pakistan imports most of its semiconductors and has no indigenous FAB unit. The country recently unveiled its semiconductor roadmap (the Pakistan National Semiconductor Plan (PNSP)) in January 2022. Under this plan, Pakistan has aimed to position itself as a potential centre for chip design and light fabrication. Although a few chip design houses are based in Pakistan, lack of skilled human capital and intellectual resources currently impede the development of top tier designing capabilities in the country. But this is precisely where a country like Pakistan can benefit from tech cooperation with China- and US-based design companies. Pakistan has surplus human capital with 25,000 engineers graduating each year in the country. With further training, these engineers could become a considerable part of the chip-designing process and could help meet the critical workforce needs in China and the US.

Similarly, testing and packaging of semiconductors could be the most significant opportunity for Pakistan to enhance its standing as a major contributor to the global supply chain of semiconductors. Testing and packaging is cheaper, suits the market trends in Pakistan and involves different skills as compared to manufacturing.

Companies such as AltaNova are already offering semiconductor testing services in Pakistan. In order to further expand testing and packaging operations, the country will need to collaborate with prominent testing and packaging companies of the world, and encourage regional companies to invest in the country.

Further, Pakistan’s diaspora working in different foreign companies could also play a key role in bringing tech expertise back home. Developing skills and capabilities in any component of semiconductor supply chain could also help solve some of the key economic issues faced by Pakistan, as the chief architect of PNSP, Dr Naveed Sherwani predicted: ‘the industry could bring up to US$4 billion in Forex to Pakistan each year if the country can get 100,000 semiconductor talent involved in the next 5 to 6 years.’

Finally, Pakistan also has the potential to become a key data resource, particularly since it is a part of China’s Digital Silk Road Initiative.  But the country has to strengthen its domestic digital infrastructure and data privacy measures. It also has to improve its regulatory environment to attract foreign investment, and promote greater cooperation with other countries in the region. By doing so, Pakistan can leverage its strengths to become a significant player in the Digital Silk Road, and contribute to the development of a more connected and digitally integrated Asia.

*

The AI race between the US and China has significant implications for South Asia, for both its socio-economic and geo-security landscape. While the ongoing competition may pose certain challenges for India and Pakistan, there are also numerous opportunities for these countries to leverage the race to their advantage. By positioning themselves as important players in the global supply chain of microchips and data resources, both nations could capitalise on the lucrative opportunities that the AI race offers. The ultimate gaol should be to create channels for global tech cooperation and hardware openness in order to avoid monopoly and possibility of tech colonialism.

However, to fully realise these goals, developing countries will need to formulate comprehensive policies and strategies that enable them to build and enhance their domestic semiconductor capabilities, attract foreign investment and foster regional cooperation in the tech sector. Countries like Pakistan and India also have to strike a balance in their foreign relations with China and the US in order to avoid being caught in any crossfire.  Ultimately, the US-China AI race will continue to shape the global technology landscape, and South Asian nations must be prepared to navigate and adapt to these changes in order to remain competitive and relevant in the years to come.

*

The views expressed here are those of the author and do not represent the views of the ‘South Asia @ LSE’ blog, the LSE South Asia Centre or the London School of Economics and Political Science. 

This blogpost may not be reposted by anyone without prior written consent of LSE South Asia Centre; please e-mail southasia@lse.ac.uk for permission.

Banner image © Hitesh Choudhary, ‘Artificial Intelligence is on its Way’, 2018, Unsplash.