Big Tech Official News (Feb 5, 2026 Issue)

Update Date: 2026/2/5

Executive Summary
On February 4, 2026, Apple integrated Claude Agent/Codex into Xcode 26.3, fully introducing "Agentic Coding" with MCP support. Anthropic released 11 OSS plugins that automate legal and analytical tasks, causing legal SaaS stocks to plummet due to the acceleration of contract reviews and similar processes. Intel welcomed Eric Demers to its GPU division to strengthen AI GPUs for data centers using the 18A process. Microsoft launched a content marketplace where publishers are compensated for content used in training and grounding, accelerating the battle for dominance in "development, operations, and data" in the age of agents.

1️⃣ Apple fully introduces "Agentic Coding" in Xcode 26.3

Source URL: Apple Official, The Verge
📝 Summary:
Apple has introduced "Agentic Coding" in its integrated development environment, Xcode 26.3. AI agents like Anthropic's Claude Agent and OpenAI's Codex can now autonomously perform code writing, editing, and configuration changes on behalf of developers. By supporting the open Model Context Protocol (MCP), it allows for the integration of third-party AI tools. This is the first instance of a company introducing third-party generative AI agents into its own ecosystem, and significant improvements in development efficiency are expected.
⚡ Strategic Significance: A move that brings a turning point to Apple's AI strategy, which has been described as conservative. The automation of the development environment could dramatically change the role of software engineers from "people who write code" to "orchestrators who direct AI agents."

2️⃣ "SaaS Apocalypse" occurs following the release of Anthropic's legal plugins

Source URL: Legal IT Insider, The Guardian
📝 Summary:
Anthropic released 11 open-source plugins that automate legal and data analysis tasks, causing the stock prices of legal research SaaS companies to crash. Hundreds of billions of dollars in market capitalization were wiped out globally, including Thomson Reuters (-18%) and Relx (-14%). The ability to complete contract reviews, NDA processing, and compliance tracking in minutes has collapsed the competitive advantage of existing professional SaaS.
🔥 Market Impact: Model providers are transforming into "owners" of the application layer. This is a historic turning point where AI has shifted from "Service as a Software" to "Software as a Service (replacement of labor)." Traditional domain-specific SaaS companies are now forced to find new points of differentiation.

3️⃣ Intel hires Eric Demers as Chief Architect of the GPU division

Source URL: The Tech Buzz, Reuters
📝 Summary:
Intel has hired a veteran who led GPU design at AMD and Qualcomm to challenge NVIDIA in the AI accelerator market for data centers. Moving away from its traditional CPU-centric approach, the company is focusing on the design and manufacturing of its own AI GPUs using the 18A process node. This is seen as a "make-or-break" move to break the market monopoly held by NVIDIA and AMD.
💪 Changes in the Competitive Landscape: By vertically integrating from design to manufacturing, there is a possibility of achieving better cost efficiency than competitors who outsource to TSMC. However, HBM memory supply shortages are expected to continue until 2028.

4️⃣ Microsoft officially announces "Publisher Content Marketplace" for AI training

Source URL: Microsoft Official, Indian Express
📝 Summary:
A hub has been built where AI developers can pay publishers directly to use high-quality "premium content" for model training and grounding. This presents a transparent economic framework to address copyright infringement issues caused by unauthorized scraping. It provides an environment where publishers can set their own usage terms and pricing, and AI developers can "sign up with a single click."
📚 Sustainability of the Content Ecosystem: The first full-scale attempt by Big Tech to form a market to make the publishing ecosystem sustainable in the AI era. It contributes to improving the accuracy of Copilot's answers and reducing hallucinations.

Comprehensive Analysis

With Apple shifting to "directed development" by welcoming external AI agents to the development site, Anthropic disrupting the legal application layer with plugins and shaking the revenue structure of SaaS, Intel re-entering the AI GPU race with vertical integration of design and manufacturing, and Microsoft establishing a legitimate distribution market for training data, the points of control for "tools, models, data, and hardware" are being reconfigured around workflows autonomously executed by agents. Victory is being determined less by speed and more by design capabilities in verifiability, accountability, and supply constraints.

Points to watch in the future

• The key will be how deep the autonomous operations (configuration changes, dependency management, PR creation, etc.) allowed by Xcode's MCP support will be, and whether audit logs and permission designs will be standardized.

• Anthropic's legal plugin is a symbol of "models becoming app owners," and the differentiation of existing SaaS will shift toward proprietary data, practical control, and accountability.

• For Intel, success will depend not only on AI GPU implementation on 18A, but also on developer experience (SDK/optimization/porting support) to compete with CUDA and the securing of HBM.

• The breadth of publisher participation, pricing design, and transparency of usage terms (training/search/summarization) in Microsoft's content marketplace will determine its reach.

• With the spread of agents, 'liability for failures,' 'copyright and data provenance,' and 'verification procedures' are becoming competitive advantages, and movements in regulation, standards, and insurance also require monitoring.