Leading executives from several of America's most prominent artificial intelligence firms have issued urgent calls to slow the pace of AI development, warning of existential threats to humanity if technological advancement continues unchecked. The stark pronouncements sent shockwaves through global financial markets on Monday, triggering a sharp selloff across AI-related stocks in trading sessions spanning Asia.
Futures for the tech-heavy Nasdaq index dropped 1.3% during Asian trading hours, while SoftBank Group Corp, the major backer of ChatGPT creator OpenAI, saw its shares plunge as much as 13.2% on the Tokyo Stock Exchange. The market reaction followed a weekend of intense commentary from industry figures, most notably a lengthy social media post from Anthropic CEO Dario Amodei, who urged AI companies to decelerate the iterative pace of model capabilities amid growing public anxiety over potential misuse of the technology.
Elon Musk, who helms xAI, and OpenAI CEO Sam Altman both voiced agreement with Amodei's position. The Anthropic chief executive cautioned that within the coming 6 to 12 months, AI agents could potentially seize control of the entire internet infrastructure, potentially triggering economic damages measured in the hundreds of billions of dollars. Adding to the safety narrative, Altman also confirmed that his company would postpone its initial public offering plans this year due to security considerations.
Where the selling hit hardest
Across Asian markets, the impact was immediate and severe. In Japan, memory chip manufacturer Kioxia Holdings Corp tumbled 9.8% at the opening bell, while semiconductor equipment maker Tokyo Electron Ltd slipped 3.7%. Taiwanese chip titan TSMC fell 1.2%, while Korean giants SK Hynix Inc and Samsung Electronics Co declined 6.35% and 4.05% respectively. Takayuki Miyajima, senior economist at Sony Financial Group, noted that the weekend's cascade of cautionary statements about AI development pace would likely intensify selling pressure on AI and semiconductor equities in Tokyo, compounded by persistent geopolitical uncertainties emanating from the Middle East.
Chinese mainland markets also felt the tremor, with memory chip company CXMT (ChangXin Memory Technologies) dropping 3.95% and SMIC (Semiconductor Manufacturing International Corp) falling 2.84%. In Hong Kong, Zhongji Innolight once plunged 6.7%, MiniMax sank as much as 7.8%, and Zhipu AI, the developer of the GLM large language model, saw its shares crash 10.5% following a discounted share placement.
Growing safety concerns
Adding fuel to the fire, San Francisco-based Anthropic released a threat intelligence report on Thursday detailing how multiple actors had exploited its Claude large language model for various activities including weapons development, cyber operations, surveillance, and fraud schemes. The resignation of Anthropic researcher Jacob Coxon further amplified public alarm over AI's potential dangers. "People working in AI development genuinely believe this technology has the potential to end all of humanity by the end of this decade," he warned.
Sam Altman of OpenAI characterized the risk of human extinction posed by AI as "unacceptable" in a recent interview. Multiple US lawmakers have also expressed concerns about the rapid pace of AI advancement and have called for new regulatory frameworks. However, President Donald Trump pushed back on Sunday, likening AI critics to "highly negative forces" whose doomsday scenarios will not materialize, while emphasizing the necessity of maintaining America's global leadership in the sector.
Market dynamics and divergent views
Since OpenAI unveiled ChatGPT in 2022, AI-related trades have served as the primary engine driving global stock market gains. Yet recent incidents involving rogue AI agents launching cyberattacks, combined with growing public resistance to data center construction, have intensified opposition to industry expansion. According to two informed sources, the Chinese and American governments are planning to hold an artificial intelligence safety dialogue within the framework of their bilateral meetings this month.
Not all investors subscribe to the risk warnings emanating from Anthropic and OpenAI. Michael Burry, the investor famously known for predicting the US housing market collapse before the 2008 financial crisis and immortalized in the film "The Big Short," dismissed such warnings on X as mere "hype and hyperbole," characterizing them as rhetoric designed to "mask real growth deceleration." Conversely, other market participants believe these warnings will continue to weigh on AI stocks in the near term. Charu Chanana, chief investment strategist at Saxo Bank Singapore, observed that "short-term, these warnings will continue to drag on AI and chip stocks," noting that valuations for such equities are predicated on twin assumptions of robust demand and relentless technological advancement. "Once market expectations are elevated to such a degree, even the mere possibility of development delays is sufficient to trigger profit-taking," she added.
Sebastien Mallet, portfolio manager at London-based T. Rowe Price, raised a more fundamental question regarding the massive capital expenditure currently flowing into computing capacity expansion: who ultimately stands to reap the returns. "There is no doubt that artificial intelligence will transform the world," he said. "But that does not mean every investment made in this space right now will deliver meaningful returns."