AI Daily News — 2026-11-02 (Mon) Evening News

AI news analysis report (last 12 hours)

Overview

``Reversing'' workforce reductions due to the introduction of AI (Gartner predicts that 30% will be rehired by 2029) has become a theme on the employment front. Anthropic revealed that 26% of the next-generation Claude development will be led by AI itself. On the other hand, xAI's Grok operating costs are said to be "unsustainable," raising questions about the profitability of the AI economy.

Anthropic / Claude trends

According to internal analysis, approximately 26% of next-generation Claude development is driven by AI, up from 1% in March. These numbers indicate that R&D automation using AI has entered the stage of actual operation. On the other hand, in terms of security, VOLTA Infostealer is reported to be using Claude, Gemini, and Kimi to rank victims by asset value, and the abuse of generated AI is taking shape. Claude Code is said to have been adjusted to prevent imitation of existing works in game generation.

OpenAI trends

Sam Altman visits the UAE and signs energy and investment agreements. Adnoc becomes the first energy company to implement OpenAI Frontier. He also made a surprise visit to South Korea to expand semiconductor partnerships. On the other hand, usage limits for ChatGPT and Codex were tightened, and funding plans were also reported to be reduced. The parents of a woman who died of a seizure after ChatGPT's advice have filed a lawsuit, raising legal risks.

Google AI trends

Announced "Gemini 4 Carbon" with early access for cyber defenders. Gemini can now create and edit PDFs and documents within chat, and has added the ability to create and share AI avatars from your face. Sundar Pichai said he was the first to run an AI prompt in space. On the other hand, it has been reported that AI chatbots are putting pressure on search revenues, and there are concerns about huge AI investments.

Trends of other AI companies

- xAI: Elon Musk admits Grok's costs are "unsustainably high" relative to revenue, with weekly caps on users. - DeepSeek: Open sourced the coding agent ``dsh-superagent'' that monitors GitHub, detects problems, and suggests solutions. - Microsoft: McGraw Hill provides educational content to Copilot. Meanwhile, eight AI commands will be removed from Windows 11. Dropbox cut about 200 jobs at the same time as Copilot integration. - Meta: Mark Zuckerberg reportedly announced a humanoid robot. - Other: Cohere's sovereign AI contract, South Korea's 2027 AI budget (21.3 trillion won, nearly doubled).

Remarks and trends of AI celebrities

- Sam Altman (OpenAI CEO): Visited UAE and South Korea. His statement that PC RAM should be used in data centers received backlash from the crypto asset industry. - Elon Musk (xAI Founder / Tesla CEO): Acquires former Spam factory and shooting range to expand Colossus 2. We predicted that orbital AI would outperform data centers. He also acknowledged Grok's cost issues. - Jensen Huang (NVIDIA CEO): Congresswoman Elizabeth Warren demands an explanation from the board and Huang to Congress over reports of spying involvement. The Chinese side denies it. Trump has expressed a desire to keep him on the Science Council. - Dario Amodei (Anthropic CEO): The headline links the 26% figure for AI-driven development and the argument that in the race against China, we should manage risk, not speed. The disappearance of the ``bottom rung'' for new graduates is also being discussed. - Sundar Pichai (Google/Alphabet CEO): Said that he had successfully executed an AI prompt in space. - Mark Zuckerberg (Meta CEO): It has been reported that there is a plan to give AI a body using humanoid robots. - Demis Hassabis (Google DeepMind CEO): Featured in connection with an article discussing "the odds are against AI doomsayers." - Ilya Sutskever, Yann LeCun, Geoffrey Hinton: I cannot confirm what the person said, so I can only link them to related articles. Hinton is associated with an article discussing the risks of sycophancy.

Breakthroughs and innovations in AI technology

Automation of AI development using AI (26% of Claude development), AI reasoning in outer space, and autonomous coding agents (DeepSeek) stand out. The Sanger Institute looks back at the acceleration of science through AI, and some believe that healthcare will be the next major battleground. Edge-oriented Physical AI SoC (Ceva/Mobilint) has also appeared.

AI Regulation/Policy/Social Impact

China announced an 18-point plan to protect workers from job losses due to AI, and also released a white paper on international AI cooperation. In the United States, ChatGPT-related death lawsuits have sparked debate over the risk of sycophancy. Suspicions of spying on China surrounding Nvidia are growing as a political issue of export controls and technology leaks. In Germany, the Digital Summit focused on AI. The pros and cons of legal research using AI are being debated in the US legal industry.

Integrated analysis: What is happening now

Employment: Some of the layoffs due to AI have turned into rehiring, and the anticipated headcount reductions are being corrected. On the other hand, there is an opposing view that the structural problem of entry-level job ladder disappearing remains and that the growth of juniors like Macquarie will accelerate. Profitability: Major companies are all facing pressure on their cost structures, including Grok's operating costs, OpenAI's usage cap and procurement reduction, and Google's search revenue squeeze. In the bond market, there are warnings (related to Hartnett) that funding for AI investment is nearing its limit, reinforcing the argument that there is a crack in the AI boom. Strength of power: OpenAI is solidifying its network of capital, energy, and semiconductor partnerships in the Middle East and South Korea, while Anthropic is demonstrating its technological presence by self-accelerating development. DeepSeek will follow up with an open source agent, and Google will gradually deploy Gemini 4, starting with defensive applications. Geopolitics: Semiconductor/espionage issues with China and sovereign AI investments in each country (South Korea, Cohere, UAE) are progressing in parallel. Risk: Attack use (VOLTA) and defense use (Gemini 4 Carbon) of AI are progressing simultaneously.

Future points of interest

Trends in the ratio of AI-led AI development and its impact on safety and supervision. 2. Usage restrictions and funding trends for OpenAI, xAI, etc., and sustainability of AI investment (including bond markets). 3. Congressional investigation into Nvidia and the future of export controls to China.

Please note that this report is based on headings only. The contents of individual articles have not been confirmed, and the statements made by each person are not substantiated beyond the headline.