Interwoven Restrictions and Conflicts: U.S. Military Contractors Accelerate Phasing Out the Claude Model

According to TechCrunch, the U.S. defense technology sector is experiencing supply chain turmoil caused by regulatory policy conflicts. Although Anthropic's Claude model is currently still used by the U.S. military for operational support, a series of overlapping and conflicting bans issued by the Trump administration have led many defense technology contractors to accelerate the abandonment of the system.

According to the latest instructions, civil agencies are required to immediately stop using Anthropic products, while the Department of Defense has been granted a six-month transition period. However, before the policies are fully implemented, the conflict between the U.S. and Iran has placed the model in a sort of "wartime paradox": on one hand, Claude is deeply integrated into Palantir's Maven system, assisting the Pentagon with real-time target identification and prioritization; on the other hand, Defense Secretary Pete Hegseth has publicly pledged to include it on the list of supply chain risk.

Robot War Future Sci-Fi (1)

As a result of these expectations, a large-scale "decoupling" trend has emerged in the defense industry. Major contractors such as Lockheed Martin have already started the process of replacing the model, and venture capital firm J2Ventures has confirmed that 10 companies in its portfolio have stopped using Claude in military scenarios.

Official Social Accounts of OpenClaw Launch, Major Domestic Large Model Manufacturers Join the Interaction

On March 3rd, the official Weibo account of the open source project OpenClaw was launched, and its first post triggered interaction from multiple domestic large model manufacturers. The project has shown remarkable performance in the global AI field, is currently strong on GitHub's trending list, and has attracted attention during MWC2026.

Jack Ma and Alibaba's Core Management Appear at Yungu School, Focusing on AI Strategy and Technological Insights

Jack Ma appeared at Hangzhou Yungu School with Alibaba's core management team, discussing opportunities and challenges in the AI era with teachers and students, signaling Alibaba and Ant Group's comprehensive commitment to AI strategy. Jack Ma emphasized that AI will have a profound impact on society, and young people need to enhance their ability to cope with technological changes. This rare gathering of senior management aims to share Alibaba's recent AI initiatives.

Alibaba TONGYI Qwen Open Source Qwen3.5 Small Model Series: Multimodal Agent Can Run on Edge Devices

The Alibaba TONGYI Qwen team has launched the Qwen3.5 small model series, including four lightweight models of 0.8B, 2B, 4B, and 9B, along with their corresponding base versions. They are based on a unified architecture, equipped with native multimodal capabilities (supporting image-text processing), with structural improvements and reinforcement learning training that can be scaled, achieving higher intelligence levels with fewer computing resources. Among them, the 0.8B and 2B models are extremely compact and fast in inference, specifically optimized for edge devices.

DeepSeek V4 is Coming Soon! A New Multimodal Model Will Disrupt the AI World

DeepSeek will release the multimodal large model V4 next week, supporting image, video, and text generation. The model is fully compatible with domestic computing power, aiming to optimize compatibility with chips manufactured in China, which is expected to promote local semiconductor demand and accelerate the integration of AI inference and domestic chips.

Trump Orders to Discontinue Anthropic, Pentagon Lists It as a Supply Chain Risk

The U.S. government and AI company Anthropic have intensified their conflict, with Trump ordering federal agencies to completely discontinue using its products within six months and exclude it from the government contractor system. The Secretary of Defense has followed up on implementing this directive.