【Krisis Iran】Microsoft mengambil langkah pertama: terus mengintegrasikan model AI Anthropic , hanya mengecualikan Departemen Pertahanan AS

Microsoft (MSFT) reportedly will continue to embed artificial intelligence technology from startup Anthropic into its customer-facing products, apart from the U.S. Department of Defense.

Earlier on Thursday (the 5th), a U.S. federal agency notified Anthropic that the company would be classified as a supply chain risk entity. Anthropic subsequently announced plans to file a lawsuit over this action.

Trump Urges Federal Agencies to Stop Using Anthropic

Last week, U.S. President Donald Trump called on federal agencies to cease using Anthropic’s models, while Defense Secretary Pete Hegseth stated that Anthropic would continue to serve the Pentagon but for no more than six months. Reports indicate that the Pentagon has used Anthropic models to assist in airstrikes against Iran.

Following the Pentagon’s actions, Microsoft became the first major company to announce it would continue collaborating with Anthropic. Some defense technology firms have instructed employees to stop using Anthropic’s Claude model and switch to alternative solutions.

A Microsoft spokesperson told foreign media that, after legal review, it was concluded that Anthropic’s products (including Claude) can continue to be provided to customers (excluding the U.S. Department of Defense) via platforms such as M365, GitHub, and Microsoft AI Foundry. The company also stated it can continue working with Anthropic on non-defense-related projects.

Microsoft Provides Technical Support to Government Agencies

Microsoft provides technical support to multiple U.S. government agencies. Microsoft 365 office software is widely used within the Department of Defense. In September, Microsoft announced the integration of Anthropic’s generative AI models with OpenAI’s models into the Microsoft 365 Copilot plugin for Microsoft 365 subscribers.

Many software engineers have adopted Anthropic’s Claude model for coding, and these models, along with OpenAI’s similar Codex model, are available through GitHub Copilot AI development services.

In November last year, Microsoft announced that Anthropic committed to investing $30 billion in Microsoft Azure cloud services, and Microsoft also agreed to invest up to $5 billion in Anthropic. However, Microsoft’s collaboration with OpenAI is even larger; as of October last year, Microsoft held $135 billion worth of shares in OpenAI, which has committed to investing $250 billion in Azure.