AI Giants Call for 'Slowdown' as Chip Stocks Plummet; Cybersecurity ETF Posts Record Gain

Multiple foreign media outlets reported that three major tech giants—Anthropic, OpenAI, and xAI—have jointly called for a slowdown in artificial intelligence (AI) model development. This triggered market concerns about weakening demand for AI hardware and cooling capital expenditures on data centers, leading to a sharp sell-off in chip, AI cloud computing, and related infrastructure stocks on Monday (14th). However, Wall Street analysts argue there are currently no clear signs of an imminent reduction in AI spending, noting that the initiative does not equate to halting model training, suggesting the market reaction may have been excessive.

Dario Amodei, CEO of Anthropic, urged the global industry to slow down the development of cutting-edge models and strengthen safety evaluations. Elon Musk, head of xAI, and Sam Altman, CEO of OpenAI, subsequently voiced their support, marking a rare consensus among the 'three AI giants' on slowing development pace.

Selling pressure concentrated heavily on the AI hardware supply chain. The iShares Semiconductor ETF (SOXX-US) plunged 5.63% on Monday, Micron (MU-US) dropped 5.25%, Marvell (MRVL-US) fell 7.32%, and NVIDIA (NVDA-US) declined 3.36%.

AI cloud computing service providers also faced significant pressure. Nebius (NBIS-US) declined 5.50%, and CoreWeave (CRWV-US) dropped 6.75%. These companies’ growth prospects are highly tied to AI model training, computational demand, and data center expansion, making them primary targets in this wave of selling.

Wall Street: Not a Halt to AI Development

Federico Manicardi and Victoria Campos from JPMorgan International Trading stated that the initiative does not call for a complete halt to AI industry development. They emphasized that the benefits AI brings to economic growth, healthcare, and democratic societies still outweigh potential tail risks.

They noted that trends in AI application usage and token consumption remain unchanged, and excessively slowing development could cause the U.S. to lose its leadership in AI. The Trump administration continues to support AI industry advancement.

Jeffrey Favuzza, equity trading analyst at Jefferies, pointed out that chip stocks were already significantly below recent highs when new AI safety concerns emerged.

Favuzza added that the statements did not clearly signal an upcoming slowdown in spending. Amodei himself explicitly stated that slowing the pace of AI capability improvements does not mean stopping model training.

From Training to Inference: Uncertainty in Data Center Demand

Bernstein analyst Madison Rezaei suggested that if AI model training truly slows, demand for remote data centers could decline, as many facilities are specifically built for model training and do not require near-instantaneous data transmission.

However, as AI inference increasingly accounts for a larger share of data center workloads, investors will closely monitor the speed at which demand shifts from model training to inference.

Rezaei noted the answer is unclear, as the market still cannot determine which AI applications will achieve widespread adoption and profitability. Agent-based AI systems typically require lower-latency infrastructure, meaning such workloads could continue to support real-time data transfer and data center demand.

Capital Flows into Cybersecurity Stocks

As chip and AI infrastructure stocks were sold off, capital rotated into software and cybersecurity firms. The Global X Cybersecurity ETF (BUG-US) surged 10.66%, recording its largest single-day gain since inception.

CrowdStrike (CRWD-US) jumped 13.85% on Monday, and Palo Alto Networks (PANW-US) rose 13.09%, reflecting market expectations that rising AI security concerns will drive enterprises to increase spending on cybersecurity and monitoring.

Monday’s market movement reflects a broader reassessment of potential winners and losers within the AI sector. Chipmakers, cloud computing providers, and data centers focused on model training face concerns over slowing capital expenditure. Meanwhile, cybersecurity and certain software firms may benefit from increased corporate investment in AI risk management and protection.