Is AI boosting economic growth?…‘Not yet’

AI technology has been rapidly gaining prominence in recent years. ChatGPT 3.5, launched in late 2022, marked a shift from AI being a future concept to a reality. While this excitement has spurred optimism about AI's potential to boost economic growth, concerns have also been raised about the potential for mass unemployment, inequality, and even a dystopian future.

Some argue that AI's rapid development could pose significant risks, calling for a slower pace of development. The question remains: what future will AI bring to humanity? This is arguably the most challenging question for our era to answer. In this article, let's examine how AI is currently impacting economic growth. Currently, the answer is a mix of 'yes' and 'no'.

On the positive side, the current AI boom is contributing to higher growth rates. In the United States, for instance, GDP growth in Q1 2026 was 2.1%, increasing to 1.5% in Q2. However, the most significant point of interest is the substantial investment in AI-related sectors. In the second quarter of 2025, Harvard professor Jason Perman noted that 92% of the United States' economic growth was due to AI-related investments, with the remaining 0.1% falling short of growth.

Despite other sectors stagnating, investments in information processing equipment, software, and technology drove the economy's growth. Similarly, in Q2 2026, the U.S. economic growth was fueled by increased investments from big tech companies, including data centers. If we calculate this investment, information equipment and software investment accounted for 69% of GDP growth.

Of course, not all investments are directly related to AI, and the trade deficit resulting from the U.S.'s import-heavy economy could negatively impact growth. Therefore, the impact of AI-related investments on U.S. economic growth is debated, with recent reports suggesting that their contribution ranges from one-third to half. In South Korea, the recent 'AI boom' is also fueled by the surge in semiconductor exports.

Q2 2025 saw a 3.7% real GDP growth increase, while the year-over-year nominal growth rate reached an impressive 26%. This was primarily due to the significant increase in semiconductor exports, up 163% YoY, resulting in a 5-fold increase in the trade surplus. Just like in the U.S., the growth in AI investment and exports can boost GDP, contributing to economic growth.

However, this growth boost is limited to countries with a strong presence in AI and semiconductor industries, and the impact of a potential AI bubble or downturn on growth should be noted. The International Monetary Fund (IMF) report also highlights that AI-related investments have experienced a rapid increase, comparable to past infrastructure investments.

However, the recent study in the Brookings Conference Conference suggests that AI infrastructure investment is growing rapidly and may become over-invested, posing a risk to the economy. Moreover, recent Economist reports indicate that the revenue and profits of AI service companies like OpenAI are rapidly increasing, but the speed of investment growth may not justify the investment.

Additionally, studies show that major tech companies, such as the 5 hyper-scale manufacturers, are investing more in capital expenditures than operating cash flow, leading to significant debt and external financing needs. The complex financial structure and leasing arrangements associated with data centers pose risks not reflected in their balance sheets.

History shows that revolutionary technologies often bring growth, but they also bring bubbles and subsequent economic crises. The current situation of AI-driven investment growth may not align with the optimistic views of AI proponents. While the optimistic scenario often discussed involves AI replacing human mental labor, allowing for more output with the same labor, the more common long-term scenario sees AI boosting labor productivity (GDP per labor hour).

This would result from AI's contribution to the 'total factor productivity growth' (TFP) that separates the increase in output from labor and capital input. Although some experimental studies show that generative AI like ChatGPT has improved workers' productivity in their respective work areas, it is unclear if the overall economic productivity is being enhanced.

Unfortunately, the answer to this question remains negative. As shown in the graph, the total factor productivity growth rate has not increased in the U.S. economy since AI became widespread in 2023. Particularly, post-pandemic, total factor productivity growth has been relatively low. This suggests that the productivity boost is not due to technological innovation but rather demand expansion and rising sales, resulting in higher productivity.

A recent Bloomberg report notes that AI infrastructure investment in the U.S. alone is projected to reach 3.6% of annual GDP by 2025-2032, outpacing past infrastructure investments. However, the sustainability of such massive AI-related investments is questionable. Recently, the Economist pointed out that while AI service companies like OpenAI have seen rapid revenue and profit growth, their investment growth has not kept pace.

Furthermore, studies indicate that major tech companies, such as the 5 hyper-scale manufacturers, are actively financing their capital expenditures through debt and external financing, posing risks not captured in their balance sheets. As history has shown, revolutionary technologies often bring growth but also economic crises. While history may repeat itself, it is essential to watch the potential economic fluctuations caused by this AI-driven investment growth.