This article was first published in the forum, The Edge Malaysia Weekly, from March 23, 2026 to March 29, 2026.
Artificial intelligence (AI) is widely seen as the next engine of economic growth. Governments and businesses around the world are investing billions of dollars in AI technology to improve efficiency, innovation, and competitiveness. But a puzzling trend is emerging. Despite rapid advances in artificial intelligence, productivity growth remains slow in many countries.
This disconnect, often referred to as the AI productivity paradox, raises important policy questions. Why has the surge in investment in AI not yet translated into broader productivity gains?
AI has the potential to transform production, innovation, and service delivery. However, the economic benefits of new technologies are rarely immediate. Even early technological revolutions, such as electrification and the Internet, took long adjustment periods before productivity gains were fully visible.
Therefore, it is important to understand and avoid AI paradoxes. The real question is not whether countries have adopted AI, but whether they have the readiness and complementary drivers, such as the training and upskilling, institutional capacity and organizational transformation, needed to turn technology investments into sustainable gross domestic product growth.
The AI productivity paradox refers to the persistent contradiction between the rapid advancement and widespread adoption of AI technologies and the relatively modest improvements observed in macroeconomic productivity indicators. Evidence from developed countries suggests that productivity growth remains subdued, despite increased investment in digital technology and automation.
Between 2005 and 2016, labor productivity growth in the United States was only about 1.3 percent per year, well below the surge recorded during the technology boom of the late 1990s. A similar pattern can be seen in the UK, where productivity growth averaged just 0.5% per year from 2010 to 2022, despite rapid digital adoption.
Although this paradox appears to be universal, it is transitory, and its severity depends less on a country’s level of development than on its readiness to coordinate the deployment of the complementary assets and AI needed to turn technological potential into real productivity gains.
Evidence shows that some countries, such as the Nordic countries, South Korea and Singapore, have been successful in translating digital adoption into productivity gains, while in others, such as the UK and Turkiye, large gaps persist between investment and outcomes.
The AI productivity paradox is often explained through Solow’s J-curve, where productivity can decline or remain stagnant during the early stages of an AI investment. However, over the long term, as businesses and economies overcome these transitional challenges, the trajectory will gradually shift toward the upward phase of the J-curve. Once organizations are restructured, employees are reskilled, and complementary innovations are fully integrated, AI technologies will begin to deliver measurable productivity gains.
These results manifest in increased operational efficiency, process optimization, cost reductions, improved decision-making, and the emergence of entirely new business models and markets.
These trends indicate that productivity gains from AI won’t happen quickly. Rather, they tend to spread gradually across companies and sectors, often requiring years of organizational and institutional adjustment before their full economic impact becomes visible.
Importantly, this paradox is not unique to any particular country. This reflects a broader transition stage in technological transformation. Economies that have successfully translated investments in AI into productivity gains typically combine technology adoption with strong complementary capabilities, such as workforce skills, digital infrastructure, and effective innovation ecosystems.
Therefore, global experience suggests that the productivity paradox is not a permanent condition. Rather, it represents a transition period in which economies gradually adjust their institutions, labor markets, and production systems to take full advantage of new technologies.
AI transformation in Malaysia: Progress amid persistent gaps
Malaysia is entering a critical phase in its digital transformation journey. Investment in digital infrastructure, intellectual property products, and data-driven technologies has steadily increased in recent years as businesses and policymakers recognize the importance of AI in enhancing economic competitiveness.
However, productivity gains remain modest. AI-related investment is increasing, but it is slower than expected to translate into tangible improvements in labor productivity. This reflects the classic feature of the AI productivity paradox, where technology investments scale rapidly but productivity gains emerge only gradually.
Overly optimistic expectations for AI also risk leading to what analysts describe as the “trough of disillusionment,” a stage where initial enthusiasm wanes when tangible economic benefits do not materialize.
Therefore, the challenge for Malaysia is not just to accelerate the adoption of AI, but to ensure that investments in AI lead to productivity improvements across industries. Achieving this requires a strong ecosystem that supports technology adoption, workforce capabilities, and organizational transformation.
Malaysia benchmark against regional leaders
A comparison with regional peers highlights Malaysia’s position in the AI productivity paradox.
Countries such as Singapore, South Korea, and Japan are showing more stable trajectories where AI-related investments continue to increase while productivity increases steadily. Singapore, for example, recorded approximately 9.2% annual growth in AI investment and 3.5% growth in productivity. South Korea recorded an investment growth rate of 7.7% and a productivity growth rate of 3.2%, while Japan recorded an investment growth rate of 1.1% and a productivity growth rate of 0.4%.
Although investment still grows faster than productivity in these economies, the gap between the two is narrowing and becoming more stable. This suggests that these countries are moving to a more mature stage of technology adoption.
In contrast, Malaysia appears to still be adjusting intensively, with productivity gains lagging behind capital investment in AI and related technologies. Empirical evidence from the AI Productivity Paradox in Malaysia Report (2025) shows that although investments in AI have a positive and statistically significant impact on labor productivity, the strength of this relationship remains weak compared to more advanced digital economies.
This suggests that while Malaysia is increasing investments in AI, the complementary conditions needed to fully translate these investments into productivity gains are still evolving.
Turn your AI investments into productivity gains
The experience of leading in the digital economy offers several important lessons. Countries that have successfully translated their investments in AI into productivity gains have many common characteristics.
First, they have a strong human capital ecosystem. Education and training systems can quickly equip workers with the digital and AI-related skills needed in modern industry.
Second, innovation capabilities play an important role. Economies with strong research and development capabilities and effective cooperation between universities and industry are in a favorable position to absorb and introduce advanced technologies.
Third, effective institutions and a coordinated industrial strategy can help reduce transition costs associated with technological change. These frameworks facilitate faster adoption of technology across companies and sectors.
Finally, a robust digital infrastructure and data ecosystem will enable artificial intelligence to be more seamlessly integrated into production processes and service delivery.
The policy implications for Malaysia are clear. Investing in AI alone will not automatically lead to increased productivity. Greater focus needs to be placed on enhancing the complementary capabilities that enable AI to create economic value.
This includes expanding workforce reskilling programs, strengthening digital infrastructure, strengthening innovation ecosystems, and improving collaboration across technology and industrial policy.
If Malaysia succeeds in aligning AI investment with skills development, innovation capacity, and institutional readiness, the country can emerge from the AI productivity trap. More importantly, we can position ourselves as a competitive digital economy in ASEAN, where artificial intelligence acts not just as a technology trend but as a real driver of productivity, growth and economic resilience.
Dr Mohamad Norjayadi Tammam is the Deputy Director-General of Malaysian Productivity Corporation (MPC), where he leads the formulation and implementation of strategic policies to increase the country’s productivity and competitiveness. This opinion piece is part of an ongoing series by The Hive exploring how private capital drives innovation and growth in Malaysia and ASEAN.
Save by subscribing to us for your print and/or digital copy.
P.S.: The Edge is also available on Apple’s App Store and Android’s Google Play.
