According to the information compiled by the AA correspondent from the analysis published by the United Nations Trade and Development Organization (UNCTAD), although many countries and companies play a role in the production of artificial intelligence technologies, the profits are collected at certain stages of the supply chain.
While raw materials such as copper used in artificial intelligence infrastructure are supplied from countries including Chile and Peru, economies such as South Korea and Taiwan stand out in semiconductor production, and Malaysia and Mexico in assembly processes.
However, the spread of the global production network to different geographies does not ensure that the economic return is shared equally between countries and companies.
According to World Trade Organization (WTO) data, technology and artificial intelligence-enabling products, including chips and servers, accounted for approximately one-sixth of global trade in goods in 2025, accounting for 42 percent of the growth in value of goods trade in the same year.
While this development revealed that the importance of artificial intelligence technologies in global trade is rapidly increasing, it also brought to the agenda discussions about how the added value created in the production chain is distributed.
UNCTAD examined the production process of the advanced NVIDIA GB200 NVL72 server system used in artificial intelligence data centers and analyzed the distribution of the economic value created among suppliers, employees and companies.
In the study carried out on a server system worth approximately 3.3 million dollars, it was possible to monitor which activities and supplier groups accounted for more than 87 percent of the total added value.
68 percent of the approximately 2.92 million dollars of added value that could be monitored within the scope of the analysis was the profit obtained after operating expenses and corporate tax.
It was stated that most of this profit went to NVIDIA, which designed the advanced chips used in the system and outsourced the production processes to other companies.
Thus, it was revealed that although the physical manufacturing in the production of artificial intelligence hardware is carried out in different countries, the companies that control the design and technology development activities receive a higher share of the economic return.
In UNCTAD's analysis, the difference between the income earned by employees in the production of the advanced artificial intelligence server system and the after-tax profit of companies attracted attention.
While 4.3 percent of the traceable added value went to employees directly involved in production, the share of non-production employees such as engineers and managers was calculated as 9.9 percent.
Thus, while the total share of employees remained at 14.2 percent, the share of profit after tax reached 68 percent.
The data in question showed that although the labor in the production of artificial intelligence hardware was distributed among different countries and fields of activity, a significant part of the economic return was concentrated in the owners of capital and technology.
Extensive supply chain in artificial intelligence production
The production of artificial intelligence servers covers a wide supply chain, from mining to semiconductors, from energy equipment to cooling systems.
However, the economic gains of the countries in this chain vary significantly depending on the nature of the production activities they undertake.
According to the research cited by UNCTAD, the total value of silicon, copper and tin used in the advanced server system examined corresponds to only 0.2 percent of the final sales price.
This situation reveals that countries that provide critical raw materials for artificial intelligence technologies receive a limited share of the economic value of the final product.
Although the growth in artificial intelligence investments has the potential to increase the mineral exports of developing countries, this increase is not expected to increase their share in the earnings from the production chain to the same extent.
On the other hand, while the assembly of advanced server systems is mainly carried out in Taiwan, Mexico, the USA and Malaysia, the returns obtained from these activities may lag behind high value-added areas such as technology design.
According to UNCTAD's assessment, although participating in global technology supply chains creates new employment and export opportunities for developing countries, the transition to higher income-generating activities is becoming increasingly difficult.
In particular, access to advanced technology, financing opportunities and geopolitical relations are among the main factors that determine the positions of countries in the artificial intelligence production chain.
In the organization's "Trade and Development Report 2026" study, it was emphasized that developing economies should strengthen their local suppliers and develop qualified workforce capacity instead of focusing only on attracting foreign investment.
However, it was stated that the production of advanced artificial intelligence components and products remains quite challenging due to high technological and financial requirements.
UNCTAD noted that countries' participation in global supply chains is an important start, but the activities they undertake in these chains are one of the main factors that determine the economic return they will obtain.
