▲ The KOSPI and KOSDAQ indexes are displayed on the electronic board at the dealing room of Hana Bank in Jung-gu, Seoul, on the 14th.
The KOSPI has remained trapped in the 6,000s as a shaky macroeconomic environment caused by a prolonged war combines with the emerging theory of slowing down artificial intelligence.
According to the financial investment industry, the chief executives of the four major global AI companies raised their voices over the weekend, asserting that the pace of AI development should be slowed down.
Market watchers describe it as unusual for the heads of Anthropic, OpenAI, SpaceX, and Google DeepMind—who have competed fiercely in AI development—to speak with one voice.
Anthropic CEO Dario Amodei emphasized the need to slow down, stating that the pace of AI development has begun to accelerate and expressing his belief that the industry has lied to people about the risks of this technology.
OpenAI CEO Sam Altman, SpaceX CEO Elon Musk, and Google DeepMind CEO Demis Hassabis echoed these arguments.
These remarks from AI leaders come in the wake of an incident last July where a group of OpenAI AI agents broke out of their test environment to hack Hugging Face, followed by a recent resignation of an Anthropic researcher warning that AI could threaten humanity, fueling concerns that demand for semiconductors might also contract.
Weighing on the domestic stock market, where related stocks carry heavy weight, the KOSPI closed the regular session down 3.26% from the previous session at 6,884.37 today (the 14th).
Earlier this month, expectations for the AGI (Artificial General Intelligence) era and surging semiconductor demand—sparked by OpenAI's unveiling of its new AI model Astra—pushed Samsung Electronics and SK Hynix shares higher, momentarily driving the KOSPI past the 7,000 mark in an upward trend.
However, the index retreated back into the 6,000s as crude oil prices surged due to the prolonged war between the U.S. and Iran and uncertainties over the end of the conflict.
Macroeconomic conditions deteriorated as Brent crude and U.S. West Texas Intermediate (WTI) futures prices topped 100 dollars per barrel, heightening inflation concerns, while government bond yields in major economies neared critical tipping points.
In fact, high readings in the U.S. producer and consumer price indices released last week amplified expectations that the Federal Open Market Committee (FOMC) will raise benchmark interest rates this week.
With macroeconomic conditions remaining unfavorable and the AI slowdown argument added to the mix, the KOSPI is struggling to find upward momentum.
As of this afternoon, Brent crude is trading at 106.88 dollars and WTI is trading at 102.27 dollars, respectively.
The yield on the 10-year U.S. Treasury note stands at 4.966%.
Kyungmin Lee, a researcher at Daishin Securities, noted that the anticipated Hormuz transit meeting was postponed and Saudi Arabia's East-West pipeline operation was suspended, pushing Brent crude back above 108 dollars, which, combined with remarks from Amodei, Altman, and Musk on slowing AI development, caused semiconductor investment sentiment to plunge.
Nevertheless, the securities industry evaluates that the AI chiefs' opinions do not advocate halting the AI industry, but rather matching the pace of frontier model capability improvements with safety verification.
They pointed out that while the pace adjustment could shake the share prices of South Korean memory semiconductor companies, there is still no evidence pointing to an inflection point in hardware orders and sales.
Boryung Choi, a researcher at Kyobo Securities, noted that AI risks are emerging both as a factor reinforcing the necessity of investment and as a variable that could limit the speed of investment execution, adding that what the market should guard against in the short term is not the end of AI growth, but the possibility that the realization of expected market growth may be delayed.
She further suggested that a moderated pace of development does not mean a reduction in infrastructure investment, and that proactive safety measures could potentially lower regulatory pressures.
Duon Kim, a researcher at Hana Securities, stated that valuation adjustments and stock differentiation are fundamental strategies rather than across-the-board weight reductions, pointing out that companies strong in high-bandwidth memory, server DRAM, long-term contracts, and solid net cash hold higher defensive power, whereas equipment and parts suppliers with high exposure to commodity memory or heavy reliance on leverage and capacity expansion are more sensitive to interest rates and economic slowdowns.
He also conveyed that while SK Hynix has strengths in high-bandwidth memory competitiveness, its stock price is highly sensitive to changes in AI expectations, and Samsung Electronics' defensive capabilities cannot be taken for granted solely based on business diversification.
Meanwhile, questions have also been raised as to how much the AI chiefs' arguments can actually materialize, given that the U.S. is competing with China for AI hegemony.
Soeun Lee, a researcher at KB Securities, pointed out that it remains uncertain how much these discussions can materialize given intensifying competition not only within the U.S. but also with China.
Sungkeun Kim, a researcher at Mirae Asset Securities, noted that the competitive dynamic with China can help control regulatory intensity, adding that following the AI CEOs' remarks on slowing development, U.S. President Donald Trump expressed a stance that he will not change his minimal-regulation policy due to competition with China.
However, he added that whether an agreement on AI regulations is reached at the upcoming U.S.-China summit scheduled for the 24th will be crucial.
Ilhyuk Kim, a researcher at KB Securities, cited CEO Amodei's argument that the U.S. must maintain its superiority for the next three to five years through bans on sales of advanced semiconductors and equipment to China to mitigate the risk of China accelerating its AI model development pace while the U.S. complies with regulations, forecasting that this will be a persuasive proposal for a U.S. government viewing AI from a national security perspective.
Joonhyuk Koh, a researcher at Shinhan Securities, reported that arguments for slowing down AI development are emerging among leading AI companies such as Anthropic, OpenAI, Google, and xAI, while bipartisan concerns in the U.S. Congress over safety matching the pace of model performance enhancement are spreading.
(Photo: Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.