AI Daily News — 2026-09-15 (Tue)

AI industry trend report (last 12 hours)

Overview

The biggest focus is on the ``AI pacing'' controversy. President Trump dismissed Dario Amodei (Anthropic CEO)'s proposal for a slowdown as a "hoax" and agreed with Jensen Huang (NVIDIA CEO) on the phone, while other celebrities such as Elon Musk and Geoffrey Hinton joined the party expressing concerns, making the camp more visible. This controversy has spilled over into actual market sentiment, with Nvidia, AMD, and Intel falling in the stock market.

Anthropic / Claude trends

The company's co-founder spoke to the BBC about the need for a mandatory "kill switch," raising concerns about AI safety. At the same time, Anthropic's ``pacing plan'' has received backlash from the White House, with Trump also criticizing Dario Amodei by name. Furthermore, the resignation of a company researcher has caused concern among CalPERS (California State Employees' Retirement System) executives, calling it a ``humanity problem.'' On the other hand, there are reports that Google has lifted the ban on all engineers using Claude, and corporate adoption is steadily expanding.

OpenAI trends

Sam Altman points out two of AI's biggest risks, while vowing not to launch an IPO in 2026. Decrypt has reported that Altman himself has mentioned the possibility of humanity losing control of AI, with notable reports linking it to Trump's statement that he "does not downplay the risks." On the business side, the company is collaborating with AI startups for the legal industry and collaborating with Box for enterprises, and the safety controversy and business expansion are going on at the same time.

Google AI trends

The product side is steadily progressing, including introducing Gemini 3.7 Flash to AI Mode search and starting the global rollout of the Windows version of Gemini (limited to basic functions). On the other hand, some articles criticize Google's new AI licensing model for streamlining lobbying activities using a "black box," raising concerns about transparency.

[Trends of other AI companies] (Meta, Microsoft, xAI, DeepSeek, etc.)

- Microsoft: It was also reported that the company has formulated a code of conduct that puts AI under human control, and also has an internal system to prevent runaway agents. On the other hand, some analysis suggests that it is slowing down the development of advanced AI. - Meta: Discontinued AI employee tracking program following internal data leak. Cathie Wood's investment trends indicate that Zuckerberg is continuing to promote "Muse." - xAI/Musk: Resolved AI competition-related lawsuit against Apple, while litigation against OpenAI continues. Musk himself emphasized that Grok 4.8's pre-training stack is "C++ written by humans." SpaceXAI is reportedly preparing a "Claude Killer equipped with Cursor" concept before the $60B deal is finalized. - DeepSeek: A viral post that talks about the "talent, power, and humanity" of engineers has become a hot topic that gives a glimpse inside the China Frontier Research Institute. The change in development policy after V4.1 is also attracting attention.

Remarks and trends of AI celebrities

- Sam Altman (OpenAI CEO): Points out the two major risks of AI, rules out an IPO in 2026, and mentions the "possibility of humanity losing control." - Dario Amodei (Anthropic CEO): Proposed slowing down global AI and called it a "duty to humanity." Receives name-calling criticism from Trump. - Elon Musk (xAI Founder/Tesla CEO): Proposed the introduction of a peer review system and agreed with the theory of slowing down AI, while dropping AI-related lawsuits against Apple and continuing lawsuits against OpenAI. - Jensen Huang (NVIDIA CEO): Aligns AI concerns with "hoax" in phone conversation with Trump. - Demis Hassabis (Google DeepMind CEO): Reportedly agreed with Anthropic's proposal to slow down global AI. - Geoffrey Hinton (AI Godfather/Nobel Laureate): Joining calls for slowing down AI, comments such as ``The AI Godfather's warnings are becoming reality'' are appearing one after another. - Yann LeCun (Meta Chief AI Scientist): In the position mentioned in the "doomer" vs "booster" conflict in "AI War". - Mark Zuckerberg (Meta CEO): While continuing to promote Muse, dealing with internal AI tracking issues became an issue.

Breakthroughs and innovations in AI technology

The establishment of a new ``Responsible AI'' research center (BBC report) is more noticeable than the technological breakthrough itself, and there have been few reports of major technological innovations in specific models and architectures in the past 12 hours. There are articles comparing the performance of DeepSeek V4 Flash, GLM-5.3, and Qwen3.8 Flash, but there is little evidence of verification information.

AI Regulation/Policy/Social Impact

While the state of Illinois is preparing a new AI law that is said to be the first in the United States, there is a debate in the EU over who has the authority to judge high-risk AI systems. Microsoft announces a code of conduct that puts AI under human control. Bloomberg Law reports that AI data center contracts on indigenous land are creating a new legal framework. The Gates Foundation has pledged $1 billion to expand access, expressing concern that AI will widen inequality.

Integrated analysis: What is happening now

The most structural change in the past 12 hours is that the ``AI pacing'' debate has taken on the appearance of a political conflict. While the Anthropic camp (Dario Amodei, Hinton, Hassabis, Musk) appealed for safety and slowdown, the Trump administration and Jensen Huang countered with the "hoax" theory, and the market reacted sensitively to this conflict, causing stocks of Nvidia, AMD, and Intel to fall. This is more than just a technology debate, and shows that AI policy is becoming embedded within the political divide in the United States. At the same time, the transition to the "governance" implementation phase is proceeding in parallel, with Microsoft's code of conduct being formulated and Anthropic's kill switch proposal, and the two wheels of regulatory debate and self-regulation are progressing at the same time. At the corporate level, technology introductions and partnerships (TotalEnergies-Mistral, Nvidia-Palantir, etc.) are steadily expanding despite the safety controversy, such as Google's lifting of the ban on hiring Claude, and a contradictory situation in which ``discursive slowdown'' and ``practical acceleration'' coexist is emerging. In terms of geopolitics, information indicating the progress of Chinese technology, such as concerns about DeepSeek's talent drain and reports on the performance of ByteDance models, continues to emerge, and the structure of the US-China AI supremacy competition remains unchanged.

Future points of interest

How will the "AI pacing" conflict between the Trump administration and the Anthropic/Musk camp be reflected in policy (regulatory legislation and White House policy)? 2. Finalization of the content of the Illinois AI law and its ripple effect on high-risk AI determination standards in other states and the EU. 3. Will Microsoft/Anthropic's voluntary governance measures (code of conduct, kill switch) become industry standard, or will they deviate from regulations?