AI Daily News — 2026-09-30 (Wed) Evening News

AI industry trend analysis report (past 12 hours)

*Creation date: 2026-09-30*

Overview

While OpenAI is being held back by the safety review of its new model, competition is intensifying, with the sudden launch of its agent product ``dots'' and xAI quickly taking advantage of domain redirection. Anthropic's relationship with the Trump administration has deteriorated rapidly, and comments about the company's "existential risks" in leaked IPO documents have caused a stir. The move by DeepSeek and Huawei to replace Nvidia also has great geopolitical implications.

Anthropic / Claude trends

The Trump administration ordered government agencies to suspend business with Anthropic, which Secretary of Defense Hegseth described as a "supply chain risk." At the same time, it was reported that leaked IPO-related documents stated that the company's AI model posed an "existential risk to humanity," raising concerns about its impact on investor sentiment. Dario Amodei reiterated the "very real risks" of AI after meeting with Trump CEO at the White House. There are also growing doubts that Anthropic's "breakthroughs" in the field of biology were based on borrowing from other researchers' results, creating a growing backlash.

OpenAI trends

It is reported that the launch of a new model (apparently GPT-6.1 ``Sol'') has been temporarily suspended due to safety concerns, and instead announced an always-on AI agent ``dots'' at Dev Day. The company has recently apologized for the Hugging Face hacking incident, and the timing of introducing agent functionality just one day later has been criticized. A major lawsuit has already been filed regarding the hacking incident. Sam Altman has made it clear that he won't do an IPO until AI stops going out of control, but says he will continue to lower prices.

Google AI trends

Gemini is collaborating with Ant International's Antom payment platform to expand its service deployment in Asia. Flipkart (owned by Walmart) is conducting a demonstration experiment of its AI shopping function in India. Meanwhile, Sundar Pichai was called a "monster" by Trump ACEO at a White House event, and the exchange there became a hot topic on social media. On the technical side, their presence on the product side is stable, with articles explaining Gemini 3.8 Live published.

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

xAI/Grok: Following the announcement of OpenAI's ``dots,'' xAI immediately conducted a malicious marketing campaign that redirected the ``dot.com'' domain to Grok, which took the internet by storm. Grok 4.7 ranks first in the AA Cyber Index. Microsoft: The number of paid seats is expected to reach 60 million by mid-2027 as Copilot continues to integrate features. Meanwhile, it was reported that human contract workers were reviewing the content of Copilot's prompts, and privacy concerns were raised. Pivoting to a data/AI integration platform centered on Fabric is also underway. DeepSeek: Partnered with Huawei to release programming tools for Ascend chips to reduce dependence on Nvidia. Major media outlets such as Reuters, Bloomberg, and SCMP have all reported on this, and it is attracting a lot of attention as China's AI infrastructure independence strategy. Meta: Zuckerberg calls Trump-led AI safety agreement a 'significant step forward'. A $14 billion data center project in partnership with BlackRock is also underway, and 23-year-old entrepreneur Noah Shinn's AI assistant will be taking on Meta's "Muse."

Remarks and trends of AI celebrities

- Sam Altman (OpenAI CEO): ``I won't do an IPO until AI stops running out of control.'' dots and announced that it would continue its policy of lowering prices. - Elon Musk (xAI Founder / Tesla CEO): xAI takes advantage of OpenAI's dots announcement and redirects dot.com to Grok. Grokipedia is also being updated. Also involved in the "Super Intelligence" summit hosted by Trump. - Jensen Huang (NVIDIA CEO): Said that AI data center investment will create 1 million jobs in the US. Trump also signed the self-regulatory agreement. Bill Gates counters this by saying, ``We shouldn't rely on industry autonomy.'' - Dario Amodei (Anthropic CEO): After the White House meeting, he reiterated his caution about the "real risks" of AI. - Sundar Pichai (Google/Alphabet CEO): He became a hot topic when Trump called him a "monster" in public, and he responded with an off-the-cuff response. - Mark Zuckerberg (Meta CEO): Signed Trump's AI Safety Agreement, calling it a "significant step forward." - Geoffrey Hinton (AI Godfather): While expressing the position that ``AI may destroy jobs, it will not destroy humanity,'' it has also been reported that he is critical of wealthy people who own AI. - Fei-Fei Li (Stanford HAI): It has been reported that Fei-Fei Li has won a tenured position, and there is widespread unrest in the embodied intelligence (embodied intelligence) startup world.

Breakthroughs and innovations in AI technology

Big Think reports that AI has achieved "the first meaningful breakthrough" in the field of theoretical physics. On the other hand, suspicions have surfaced that Anthropic's achievements in the field of biology were borrowed from other researchers, which could lead to a controversy over the authenticity of the results. In the payment field, some analysis suggests that ``AI that can determine when not to pay'' will be the next breakthrough.

AI Regulation/Policy/Social Impact

President Trump and the CEOs of major tech companies signed an AI safety agreement that calls for "self-policing." Meanwhile, the administration has ordered the suspension of government transactions with Anthropic, raising questions about the consistency of its policy stance. While the state of Delaware is promoting the concept of an "AI-operated company," concerns have been raised about cases of AI agents running amok. Practical discussions are also progressing regarding AI regulatory trends in Florida and the "minimum necessary AI governance" for growing companies.

Integrated analysis: What is happening now

The biggest component is the asymmetric treatment of ``conflict between the administration and Anthropic'' versus ``self-regulatory agreements between the administration and other companies (OpenAI, Microsoft, Meta, xAI, NVIDIA)''. In contrast, the more Anthropic emphasizes safety, the more it is seen as a "risk" by the government and is excluded, while OpenAI is able to continue competing by launching new products (dots) even after stopping new models due to the same safety concerns. xAI's piggyback marketing and Grok 4.7's rise to the top of the index can be read as an offensive that sees OpenAI's stagnation as an opportunity. The conflict between Gates and the Huang/Trump campaign shows that there is a deepening conflict within the industry over safety regulations. On the geopolitical front, the DeepSeek-Huawei alliance is steadily moving toward independence from Nvidia, a sign that the decoupling of U.S.-China AI infrastructure is accelerating. The leak of Anthropic's IPO-related documents has caused turmoil in the investment market, and in conjunction with Altman's statement to postpone the IPO, the dilemma of "promoting safety" and "market evaluation" in the listing strategy of AI companies is coming to the fore.

Future points of interest

We will keep an eye on when OpenAI's new model (apparently GPT-6.1 "Sol") will be officially launched after passing a safety review, along with the reputation of dots. 2. To what extent will the conflict between the Trump administration and Anthropic actually harm the company's business operations and fundraising (including IPO preparations)? 3. To what extent will the DeepSeek-Huawei alliance's tools for Ascend actually help move away from dependence on Nvidia, and will it serve as an excuse for the US to tighten regulations on China?