(Jan 8): The 20th-century Cold War was defined by nuclear weapons. Today, a new cold war is being waged, with artificial intelligence (AI) as the ultimate weapon. The stakes are arguably even higher, as nations compete for control of a technology that promises to redefine our world.

The US is aggressively pursuing AI dominance. Project Stargate, a massive private-sector initiative worth US$500 billion (RM2.22 trillion) aims to solidify America’s dominance in AI development and infrastructure. Announced by President Donald Trump, Stargate is a private-sector-led endeavour led by tech giants like OpenAI chief executive officer Sam Altman, and Oracle chairman Larry Ellison and backed by SoftBank CEO Masayoshi Son. It aims to build powerful data centres in Texas, centralising AI infrastructure and serving as the nerve centre for AI-powered applications like ChatGPT and beyond.

Further reinforcing this push, the CHIPS and Science Act provides substantial funding for semiconductor development — a critical component of AI advancement. Additionally, the US has strategically imposed export controls to restrict China and other competing nations from accessing advanced AI chips and manufacturing equipment, aiming to curb their progress.

China is the US' primary rival in the AI arena. Its New Generation Artificial Intelligence Development Plan sets an ambitious goal of achieving global AI leadership by 2030. Heavily backed by state-controlled enterprises, China is investing extensively in AI research, chip manufacturing, and surveillance technology.

DeepSeek, a Chinese AI lab, showcased its innovation by developing a powerful open-source AI model for under US$6 million, rivalling systems like GPT-4 and Llama. The DeepSeek-R1 model is trained at a significantly lower cost and requires a tenth of the computing power of comparable AI models such as OpenAI's GPT-4o. While the model isn't fully open-source due to the unavailability of training data, the release of "open-weight" models is a significant step. DeepSeek's AI Assistant, powered by DeepSeek-V3, has overtaken rival ChatGPT to become the top-rated free application available on Apple's App Store in the US. Alibaba's Qwen 2.5 AI model, launched recently, has outperformed even DeepSeek V3 in benchmark tests.

The back-to-back announcements of advanced AI models from Chinese companies have raised some questions in India too. What is stopping it from developing its own foundational AI model? This was partly answered by IT Minister Ashwini Vaishnaw in a recent press conference. India's investment in over 18,000 high-end graphic processing units (GPUs) demonstrates its commitment to building robust computing facilities, essential for AI applications, model training, and algorithm development. Compared to global model computation costs of US$2.50 to US$3 per hour, India’s AI Model computation will cost less than 100 rupees (US$1.15 or RM5.06) per hour after a 40% government subsidy, significantly lowering the entry barrier for AI development.

Additionally, India is hosting DeepSeek on Indian servers after rigorous security protocol checks, allowing developers, coders, and researchers to benefit from its open-source code. The US$1.25 billion IndiaAI mission will further fuel AI startups and enhance the AI ecosystem.

While the EU may not be at the forefront of raw technological power, it is taking a leading role in ensuring AI aligns with human rights and ethical standards. The Artificial Intelligence Act seeks to regulate AI technologies, prioritising responsible and ethical implementation.

As Asean chair in 2025, Malaysia is positioning itself as a key player in the region's AI development. With a focus on becoming a digital innovation hub, Malaysia has launched the National Artificial Intelligence Office (NAIO) to guide its AI strategy. The NAIO's priorities include developing a code of ethics, a regulatory framework, and a long-term AI action plan. Budget 2025 provides further support through incentives for high-tech companies, R&D, and talent development.

The Malaysia Digital Economy Corp (MDEC) has successfully onboarded 140 AI solution providers, generating RM1 billion in revenue. AI is projected to contribute significantly to Malaysia's economy, with expectations of generating approximately US$115 billion in productive capacity by 2030.

To compete in the global AI race, Malaysia must prioritise several key areas.

1. Bolstering AI research: Malaysia needs a significantly stronger AI research ecosystem. While some initiatives exist, the funding gap compared to leading nations is vast. Top Chinese universities, for example, receive immense research funding, dwarfing that of many countries' entire academic communities. Bridging this gap requires a concerted, long-term government-industry-academia partnership focused on increasing funding for fundamental AI research, fostering industry-academia collaboration, and attracting top AI talent.

2. Closing the digital talent gap: The US, China, and India have a considerable advantage due to their readily available pool of digital talent. Despite efforts by MDEC and TalentCorp, Malaysia faces a talent shortage, especially in data science and AI. This gap is widening. A two-pronged approach is necessary: In the short term, facilitating the import of international experts to address immediate needs. Simultaneously, a long-term strategy is crucial, building a robust talent pipeline through deep industry-academia collaboration, investing in education and training to equip Malaysians with advanced AI expertise.

3. Democratising compute power: Malaysia should establish an affordable common compute facility, equipped with substantial GPU capacity, to support the development of a secure, indigenous AI model. This shared resource lowers the barrier to entry for students, researchers, and startups. The ultimate goal is to provide cloud-based AI services to a broad range of users, including academia, micro, small, and medium enterprises (MSMEs), startups, research institutions, government agencies, and other approved entities, fostering widespread AI adoption.

4. Leading in AI safety: As Asean chair in 2025, Malaysia can lead the region in responsible AI development by establishing an Asean AI Safety Institute. Operating on a hub-and-spoke model, the institute would collaborate with regional research institutions, academia, and the private sector. This collaborative approach ensures safe and ethical AI development, solidifying Malaysia's vision as a hub for digital innovation and ethical AI in Asean.

The AI race is a battle for influence and power. AI has the potential to reshape economies, military strategies, and even global narratives. Unlike the nuclear arms race, there's no clear end. The crucial question is not simply who wins, but how this competition will shape our future. Will it inspire unprecedented innovation and collaboration, or will it deepen existing divides? The AI race is happening now, and its outcome will define the 21st century.

Akhil Gupta is the group CEO of a high-growth conglomerate specialising in IT infrastructure, AI, and talent solutions.