1. The Federal Purge: U.S. Treasury and State Dept. Exit Anthropic

In an unprecedented move following President Trump's Friday directive, major U.S. federal agencies have officially terminated their contracts with Anthropic today.

• The Treasury Exit: Secretary Scott Bessent confirmed that the Treasury Department has ended all use of the Claude platform, stating, "Under President Trump, no private company will ever dictate the terms of our national security."

• The Shift to OpenAI: The U.S. State Department has already migrated its internal "StateChat" from Anthropic to OpenAI’s GPT-4.1.

• The "Pariah" Status: Analysts note that by designating Anthropic a "Supply Chain Risk," the U.S. government is treating a domestic AI lab with the same level of restriction previously reserved for foreign adversaries like Huawei.

2. Qualcomm & Cristiano Amon: Architecting 6G for the AI Era

At the MWC opening keynote this morning, Qualcomm CEO Cristiano Amon unveiled the roadmap for 6G, defining it not as a speed upgrade, but as an "Intelligent Fabric."

• From Carrying to Being: Amon argued that while 5G carried intelligence, 6G will be intelligence. The network will use "Collaborative Compute" to let your glasses, phone, and car share processing power in real-time.

• Agentic RAN: Qualcomm launched its Agentic Radio Access Network (RAN) management service, which uses AI agents to autonomously heal and optimize cell towers, preparing for a world where billions of autonomous devices are constantly "reasoning" on the go.

3. HONOR’s "Robot Phone" and Humanoid Duo

While other brands are showing concepts, HONOR made waves today by bringing its Robot Phone and a companion Humanoid Robot to the show floor for live demonstrations.

• The Species Shift: HONOR CEO James Li described the Robot Phone as a "new species" of device that integrates motion and spatial awareness. The phone can physically pivot on a gimbal-like arm to track your face during calls or "look" at objects you point to.

• The Alpha Plan: This is part of HONOR's "ALPHA PLAN," which aims to integrate "Embodied AI" into daily consumer electronics, moving the assistant out of the screen and into the room with you.

4. Samsung’s "Privacy Display" and Z-TriFold

Samsung’s booth at Fira Gran Via remained the most crowded today as they demoed the Galaxy S26 Ultra’s new hardware-level privacy features.

• The Privacy Display: Using a new pixel-blocking technology, the S26 Ultra screen becomes unreadable to anyone not looking directly at it—a direct response to the "visual hacking" concerns of 2026.

• The Tri-Fold Reality: Samsung also showcased the Galaxy Z TriFold, which expands into a nearly 11-inch tablet. It features "Photo Assist," where users can add or modify complex elements in photos using simple text prompts, powered by an on-device agent that understands the intent of the edit.

Prompt Tip of the Day: The "Agentic Architect" — Personal Security Auditor

With the U.S. government labeling software as "Supply Chain Risks," it's time for individuals to audit their own digital "supply chains." Today’s prompt builds a proactive auditor for your personal data.

The Prompt:

"Act as a professional Chief AI Architect specializing in Cybersecurity. I want to build a 'Digital Supply Chain Auditor' to monitor the apps and services I use. Please structure a framework for this agent that includes:

• Permissions Module: Instructions for the agent to analyze a list of my installed apps [insert list] and flag any that request 'High Risk' permissions (e.g., microphone or location) that don't match their primary function.

• Terms of Service (ToS) Module: Instructions for the agent to scan the privacy policies of my top 5 most-used apps and highlight any clauses that allow my data to be used for 'Third-Party AI Training' without an opt-out.

• Breach Monitoring Module: A rule for the agent to check 'Have I Been Pwned' or similar databases for any of my registered emails once a week and alert me immediately of new leaks.

• Action Workflow: A step-by-step process where the agent provides me with a 'Security Health Score' (1-10) and a list of 3 apps I should consider deleting or restricting.

For each point, provide clear, step-by-step rules that would allow an AI agent to operate as a professional and persistent security watchdog."