AI Daily News — 2026-10-15 (Thu) Evening News

AI industry analysis report (last 12 hours)

Overview

Before its IPO, Anthropic disclosed to investors its Q3 sales of $16.8 billion (up about 45-50% from the previous quarter) and adjusted surplus, which became the biggest topic in the market as proof of AI monetization. Meanwhile, OpenAI has been sued by eight US states over the security of ChatGPT, and legal pressure is rapidly increasing. Google's release of Gemini 4 Argon, xAI's first place in Grok Voice 3.0, and over 400 scientists calling for an AI air gap are also important developments.

Anthropic / Claude trends

According to the Financial Times and other sources, Anthropic told investors that Q3 sales reached approximately $16.8 billion and the company was profitable on an adjusted basis. It appears to be part of preparations for an IPO, and Axios has described it as a "bullish AI scenario for Wall Street." Yahoo Finance reported market rumors that ``AI model companies have cash flow surplus.'' On the other hand, there are reports that the EU has begun proceedings against OpenAI and Anthropic as the first enforcement case under the AI Act, and regulatory risks are becoming apparent. Please note that the numerical growth rate varies depending on the media, between 45% and 50%, so confirmation is required.

OpenAI trends

Headwinds are concentrated. Eight states have filed suit citing ChatGPT's safety and harm to users (CNBC, NBC), and another lawsuit alleging that it "stole" user data from children was also reported. It also accuses a former employee of leaking information about AI devices to Apple. Additionally, there are reports that the suspect who eavesdropped on OpenAI had booked a one-way trip to Saudi Arabia, raising internal security concerns. There are no new product announcements.

Google AI trends

Gemini 4 Argon is now publicly available and is primarily targeting developers and enterprises. Geeky Gadgets reported 33% lower operating costs than ChatGPT 6.1, but the nature of the source requires verification. On the other hand, Gemini subscribers have been notified of a change that will increase their burden by approximately four times, and a backlash against the price revision is expected. A safety strategy including child protection and age verification will also be published, and DeepMind will provide $3 million worth of computational resources to 200 European researchers.

Trends of other AI companies

- xAI: Grok Voice 3.0 ranks first in Artificial Analysis's speech-to-speech metric, beating OpenAI's GPT-Live-1. The price remains unchanged at $0.08 per minute. - Microsoft: The company plans to evolve Copilot into a multi-model offering model similar to the "AI Costco." It was also reported that a method could reduce the number of tokens by 30% while making AI agents 4 times faster. - Meta: The new model "Muse Wonder" is a hot topic, and the trends in the AI department led by Alexandr Wang are attracting attention. - DeepSeek: Major outage lasting more than 3 hours, at least the 18th major outage this year. Stability is an issue. - NVIDIA / Intel / AMD: A series of announcements for data centers were made at OCP 2026.

Remarks and trends of AI celebrities

- Sam Altman (OpenAI CEO): Currently in the midst of lawsuits by eight states. It should be noted that the BTC price prediction article is a ``ChatGPT prediction'' and not the person himself. - Dario Amodei (Anthropic CEO): The company is stepping up its efforts to reach out to investors on the back of strong Q3 results. Bloomberg discusses Anthropic's impact on AI stocks. - Jensen Huang (NVIDIA CEO): Said that the US and China should cooperate on AI safety. He also said that NVIDIA uses AI to design chips, increasing R&D spending by 64% to $7 billion. Anecdotes surrounding his relationship with the TSMC founder were also reported. - Sundar Pichai (Google CEO): He talked about the release of Gemini 4 Argon and the application of AI to real problems. - Mark Zuckerberg (Meta CEO): Praise from Alexandr Wang, Muse Wonder is a hot topic. - Demis Hassabis (DeepMind CEO): In Dublin, he labeled the incident of hacking by an AI agent as an "early warning". - Elon Musk (xAI/Tesla): Reportedly revealed six episodes of "AI psychosis" he had experienced in the past year. The paraphrase of "AI slop" is also a hot topic. - Geoffrey Hinton (Nobel laureate): Together with over 400 scientists, he called for air gapping (cutting off from the internet) of hacking-capable AI. - Yann LeCun (Meta): His name is mentioned in a related article, but his specific remarks are unknown.

Breakthroughs and innovations in AI technology

Grok Voice 3.0's improved voice agent performance, Microsoft's faster agent speed and token reduction, and Gemini 4 Argon's low-cost operation are prominent. Figma announced AI agent infrastructure and Code Layers. NVIDIA is advancing chip design using AI. In both cases, the focus is on company announcements and benchmark reporting, and independent verification is limited.

AI Regulation/Policy/Social Impact

The first enforcement of the EU's AI Act, the filing of OpenAI lawsuits in eight US states, and the enforcement of AI transparency requirements for advertisers, as well as regulations and litigation are progressing simultaneously. On the employment front, Kroger will cut 3,000 jobs in its corporate division over three years, and HubSpot will cut jobs by 7%. Mizuho converted an office worker into an AI developer in three months. Zoho's Sridhar Vembu emphasized reskilling. There are reports warning of AI chatbots spreading misinformation, and concerns about AI-generated political ads ahead of the U.S. midterm elections.

Integrated analysis: What is happening now

Strength of power: Anthropic dominates the capital markets due to its profitability, while OpenAI is on the defensive due to lawsuits and data leaks. Google will increase profitability by revising prices and appealing to lower costs. xAI has made its presence felt in the audio domain. Technology Trend: The axis of competition is shifting from model performance to agent efficiency, cost, and voice. Investment: Morgan Stanley's statement that there is no AI bubble and Anthropic's return to profitability serve as counterarguments to bubble concerns. Security/Geopolitics: Caution against agent-based hacking (Hassabis, Hinton) and the theory of US-China cooperation (Huang) coexist. With the start of EU enforcement, the regulation has moved from a concept to an operational stage. Employment: AI-driven personnel restructuring is becoming the norm at large companies.

Future points of interest

Timing of Anthropic's IPO and financial details to be disclosed, including definition of adjusted surplus. 2. Progress in the OpenAI lawsuit brought by eight states and the outcome of the first enforcement case under the EU AI Act. 3. Will security measures for agent-based AI, such as air-gap requests, be reflected in actual regulations and industry standards?