Top Stories This Morning: OpenAI May Slow Down Operations; “China’s Nvidia” Soars 206%

OpenAI has slowed down the development of its most advanced AI systems: Sam Altman told employees that the company might scale back its efforts in coordination with other labs until the industry establishes common safety standards. Anthropic detected attempts to use Claude for research potentially related to the development of biological weapons and blocked the accounts in question. Read about these and other topics in our roundup of key events as of the morning of September 11.

OpenAI is considering slowing down the development of its most advanced artificial intelligence systems, CEO Sam Altman told employees, according to Bloomberg. He said the company could slow down its pace in coordination with other AI labs, although not all of them may agree. OpenAI’s chief scientist, Jakub Pachocki, also called on companies to coordinate a slowdown in development until common safety standards are established.

Concerns within the industry have intensified following a series of incidents involving AI and public statements by researchers. For example, in July, more than 1,000 employees of AI companies signed a petition calling for the creation of a mechanism to slow the pace of the technology’s development.

Shares of Enflame, a company developing Chinese equivalents of Nvidia chips, rose 206% after its debut on the Shanghai Stock Exchange, according to CNBC. Demand for the IPO was enormous: retail investor orders exceeded the number of shares initially available to them by more than 6,000 times, the network reports. The Tencent-backed company is the latest of four major Chinese AI chip-developing startups to go public.

Investors expect local manufacturers to be able to replace Nvidia in the Chinese market. Enflame plans to use the funds raised to develop fifth- and sixth-generation AI chips. In 2025, its revenue rose to 990 million yuan ($147 million) from 722 million yuan a year earlier, though the company remains unprofitable.

Anthropic stated that it had identified and stopped large-scale attempts by Chinese AI companies Alibaba, Moonshot, and DeepSeek to use Claude’s responses to train their own models without permission, according to CNBC. The company called this practice “illegal distillation.” According to Anthropic, the largest campaign was linked to Alibaba: from May through July, more than 151 million interactions with Claude were recorded.

According to Anthropic, Moonshot redirected some of Kimi users’ queries to Claude and used the responses to train its own models—the company counted more than 23 million such interactions between May and July alone. According to the report, DeepSeek carried out more than 12 million such operations over a two-week period in July. Some of the queries forwarded to Claude contained confidential information about users and companies.

Oracle's cloud infrastructure revenue rose 121% to $7.4 billion in the first fiscal quarter, exceeding analysts' forecast of $7.19 billion, according to Bloomberg. The company’s total revenue increased by 30% to $19.3 billion, and adjusted earnings per share came in at $1.92, compared with the expected $1.75. Oracle’s stock rose by about 5% in after-hours trading.

Oracle is actively expanding its data centers for OpenAI and other clients: capital expenditures for the quarter reached $28.5 billion, and the value of new AI cloud contracts exceeded $30 billion, according to the agency. The value of future contract obligations rose to $664 billion, compared with a forecast of $641.8 billion.

Anthropic stated that it had suspended several accounts that were using Claude for research potentially related to the development of biological weapons, CNN reported, citing a statement from the company. Over the course of 30 days, the company identified approximately 35 suspicious research projects, including work on avian influenza, smallpox viruses, toxins, and gain-of-function research. However, Anthropic acknowledged that it cannot confirm that the users had malicious intent.

The company identified biological misuse as one of the most serious risks posed by AI. According to its data, users have attempted to circumvent restrictions and conceal the purposes of their research. Anthropic stated that the new models have been equipped with stricter safeguards; however, as AI capabilities continue to grow, even more robust security systems will be required.

— Japan's broad-market Topix index fell 0.8%, while the Nikkei 225 fell 2.3%.

— Hong Kong's Hang Seng Index fell 0.6%, while mainland China's CSI 300 Index dropped 1.3%.

— In South Korea, the Kospi fell 2.3%, and the Kosdaq fell 2.1%.

— Australia's S&P/ASX 200 was down 1%.

— Nasdaq 100 futures were virtually unchanged, while S&P 500 futures rose 0.2% and Dow Jones Industrial Average futures rose 0.3%.

This article was AI-translated and verified by a human editor