Claude Code : L'actualité EXPLOSIVE de la semaine !

This week marked a significant shift for Anthropic and its AI model, Claude, driven by major political and technical developments. The most striking news is a historic decision by the U.S. government. President Donald Trump has ordered all federal agencies to stop using Anthropic’s technologies due to national security concerns. The Secretary of Defense labeled the company a "supply chain risk," following a breakdown in negotiations with the Pentagon. Specifically, Anthropic refused to waive ethical safeguards regarding mass surveillance and fully autonomous systems that operate without human supervision. While agencies like the Treasury have halted use immediately, the Pentagon will allow a transition period for existing contracts to reach their legal terms. In the wake of this rupture, OpenAI stepped in, securing a deal with the Department of Defense to provide models for classified environments. While Sam Altman noted the speed of the agreement, OpenAI asserts that their contract still includes protections against domestic mass surveillance and unsupervised autonomous weaponry, with ongoing adjustments to clarify these guarantees. This highlights a growing government reliance on American AI providers other than Anthropic. On the product side, Anthropic launched "Remote Control" for Claude Code. This feature allows users to start a session on their computer and control it via a phone or tablet. Crucially, the code and environment remain local to the user's machine rather than passing through the cloud. Additionally, Claude Code now integrates Language Server Protocol (LSP), making tasks like finding function definitions up to 900 times faster and allowing for real-time error detection. To support the developer community, Anthropic introduced the "Claude for Open Source" program, offering six months of Claude Max to eligible projects. The evolution of these tools is already changing the profession; a former head of AI at Tesla recently noted that he now spends more time collaborating with AI agents than writing code line-by-line. To help users navigate these changes, a new comprehensive "cheat sheet" for Claude Code has been released, providing a quick reference for all essential commands and features. These updates suggest that while Anthropic faces political hurdles, its technical integration into the developer workflow continues to accelerate.