マイクロソフト、ナデラCEOがAIモデルを「商品」と呼んだ後、スーパーインテリジェンス追求のためAI部門を再編
本文の状態
日本語全文を表示中
詳細モードで約1分の本文を読めます。
同じ出来事の情報源
6媒体で確認
AI Business · The Decoder · LayerX Tech Blog · InfoQ · GitHub Changelog · TechCrunch AI
各社の報じ方を比較 ↓マイクロソフトは、自社AIモデルの開発強化とスーパーインテリジェンス追求に向けてAI部門を再編する。これは同社が以前「AIモデルは商品」と発言した方針からの転換となる。
Source Article
元記事を日本語で読む
本文に関係しない購読案内、埋め込み通知、サイト内プロモーションは除いています。

マイクロソフトは、自社のAIモデル開発を強化し、究極的にはスーパーインテリジェンスの実現を目指すため、AI部門を再編成しています。これは、同社が従来表明してきた方針からの大きな転換です。
本記事「Microsoft restructures AI division to chase superintelligence after Nadella once called AI models a commodity」は、The Decoderに最初に掲載されました。
原文を表示
Microsoft is restructuring its AI division to double down on its own AI models, all the way up to superintelligence. That's a notable shift from what the company used to say.
The Copilot teams for commercial and consumer customers are being merged into a single organization with four areas: Copilot experience, Copilot platform, Microsoft 365 apps, and AI models. Jacob Andreou, who previously ran product and growth at Microsoft AI and before that at Snap, will step in as Executive Vice President of Copilot Product Experience, reporting directly to CEO Satya Nadella. A new Copilot leadership team will handle strategy and product coordination across the board.
Suleyman shifts focus to superintelligence
Mustafa Suleyman, head of Microsoft AI, is going all in on building extremely capable AI models and chasing "superintelligence." According to Suleyman, these models will improve products across the company over the next five years while bringing down the direct costs of running large-scale AI services.
Nadella says progress on AI models will matter more than ever for Microsoft over the next decade, since they underpin Copilot and, by extension, much of the company's product lineup. Nadella previously argued that AI models would become commodities; going from there to superintelligence is a striking change in tone. His latest comments suggest the company now sees proprietary model development as a strategic priority.
We are doubling down on our superintelligence mission with the talent and compute to build models that have real product impact, in terms of evals, COGS reduction, as well as advancing the frontier when it comes to meeting enterprise needs and achieving the next set of research breakthroughs.
Satya Nadella
So far, Microsoft's homegrown AI models haven't been much of a factor in the race. The company unveiled large models from its internal "MAI" series for the first time in August 2025, but they fell short of the competition.
AI model providers want to be platform layers, and that's a problem for Microsoft
There's likely another factor driving this move: large model providers increasingly aim to position their services as a kind of "AI operating system," a layer that takes over and standardizes more and more work inside apps, including Windows. Anthropic is already pushing this kind of platform play with Claude Cowork. Microsoft moved fast and shipped its own version of the concept with "Copilot Cowork," adapting Anthropic's technology.
But that creates a real dependency risk. If an outside model controls the processes inside your software, you're handing over control of a critical layer. In a worst-case scenario, Microsoft could find itself unable to act independently on a core piece of its own software stack.
Nadella has been openly critical of Copilot's current capabilities internally, telling engineering managers that integrations connecting Copilot with Gmail and Outlook "don't really work" and are "not smart." He reportedly took an unusually hands-on role in fixing these issues, posting detailed critiques in internal Teams channels and grilling engineers in weekly meetings.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
Subscribe now
同じ出来事を6媒体で確認
同じ出来事を扱う別媒体の記事です。見出しと公開時刻を比較できます。
- AI Business新たなMicrosoft AIエージェントが企業の近代化を支援
- LayerX Tech BlogPython製ETL「dlt」を採用した経緯 - Azure Cosmos DB for PostgreSQLとContainer App Jobによる効率的なデータレイク構築
- InfoQ.NET 11 Preview 2が性能向上、マッピング改善、ネイティブOpenTelemetryサポートを提供
- GitHub ChangelogCopilot利用メトリクスで自動モデル選択が実際のモデル名に解決されるようになりました
- TechCrunch AIマイクロソフト、Windows上のCopilot AIの一部機能を削減
関連記事
News to Guide
ニュースの次に確認する
発表内容を、現在の料金や仕様と照らし合わせられる関連ガイドです。
今日のまとめ
AIデイリーブリーフで今日の重要ニュースをまとめ読み