Microsoft restricts Claude Fable use as OpenAI pursues $1 trillion IPO

Microsoft Corp. has restricted employees from using Anthropic's Claude Fable AI product, escalating the AI rivalry as OpenAI — backed by Microsoft — pursues a $1 trillion IPO valuation.

The internal policy was reported by Chinese financial media outlet Cailianshe, which did not disclose Microsoft's specific rationale. The restriction comes as competition between Microsoft-backed OpenAI and Anthropic intensifies across enterprise and consumer AI markets. Microsoft, which has invested more than $13 billion in OpenAI since 2019, has not commented on the report.

OpenAI confidentially filed for an initial public offering with the Securities and Exchange Commission, joining Anthropic in a push toward public markets. The ChatGPT maker is seeking up to a $1 trillion valuation, which would rank it among the most valuable companies ever to go public. OpenAI is also planning a major product upgrade to boost revenue as it battles Anthropic for market share in the generative AI sector, where corporate spending on AI software and services is projected to exceed $100 billion annually by 2027.

The move by Microsoft highlights the growing friction between Big Tech's AI investments and the independence of the tools they back. Microsoft's deep integration of OpenAI's models across Azure cloud services, Microsoft 365 Copilot and GitHub Copilot creates a competitive incentive to limit employee exposure to rival AI platforms. Anthropic's Claude Fable, a newer model in its lineup, has gained traction among developers for its safety alignment features and 200,000-token context window — capabilities that compete directly with OpenAI's GPT-5 and GPT-4o offerings. Claude has been particularly popular in enterprise settings where compliance and content moderation are priorities, eating into OpenAI's early lead in corporate deployments.

The restriction raises questions about how large technology companies will manage internal use of competing AI tools as the market matures. Anthropic, which has received more than $7 billion in combined backing from Google and Amazon.com Inc., has positioned itself as a safety-focused alternative to OpenAI. The company's Claude family of models has consistently ranked near the top of industry benchmarks including MMLU and HumanEval, often trading places with OpenAI's GPT models depending on the specific evaluation task.

With OpenAI targeting a $1 trillion IPO and both companies racing to release more capable models, corporate policies that limit access to rival AI products could become more common across the industry. For investors, the episode shows the high stakes of the AI platform war: whichever platform captures enterprise workflows first could lock in billions in recurring revenue. The restriction also puts a spotlight on Microsoft's dual role as both OpenAI's largest investor and a dominant enterprise software provider whose products increasingly rely on AI capabilities — a tension that could draw scrutiny from regulators as the AI market consolidates.

This article is for informational purposes only and does not constitute investment advice.