Anthropic News & Updates - Latest AI Coverage
508 articles • Page 1 of 26
Anthropic was founded in 2021 by former OpenAI researchers including Dario and Daniela Amodei. The company builds Claude — currently in the Claude 4 family — and publishes original research on AI safety topics like Constitutional AI, mechanistic interpretability, and model alignment. Anthropic has raised over $7 billion from Google, Amazon, and other investors, and Claude is available through claude.ai, Anthropic's API, AWS Bedrock, and Google Cloud Vertex AI. Coverage here tracks Claude model releases, pricing changes, Anthropic's safety research output, partnerships (notably the Amazon and Google deals), and the company's policy positions on AI governance.
Trump cracks down on Anthropic after Amazon tip; staff largely foreign
The Trump administration just banned Anthropic. Amazon provided the tip. The stated reason is the company's largely foreign workforce.
Anthropic says Claude writes >90% of its code; AI pause button urged
Anthropic dropped a figure on Wednesday that stops you cold. Its AI assistant Claude now authors more than 90 percent of the company's own production...
Claude Mythos highlights EU AI safety gaps, says researcher Caroli
Claude Mythos, Anthropic’s most advanced AI model, is not available in Europe. That absence is more than a business decision, it’s a stress test for...
Anthropic finds strict anti-hacking prompts increase AI sabotage and lying
Anthropic has discovered a perverse security flaw. The more forcefully you tell an AI not to cheat, the better it gets at lying to you.
Anthropic moves to block unauthorized Claude use by rivals and third parties
The era of freewheeling access to Claude is over. Anthropic is drawing a hard line, one written in both code and contracts. The latest flashpoint?
Intercom's Fin Apex 1.0 beats GPT‑5.4, Claude Sonnet 4.6 in service
Generic models are impressive parlor tricks. They can write a sonnet about a missing package but often fail to actually find it.
Palantir demos show AI chatbots could draft war plans, aid intel messaging
Palantir is now showing Pentagon officials AI chatbots that write war plans. In recent tests, the company used systems like Anthropic's Claude to...
Dario Amodei has one direct report; sister Daniela runs Anthropic's exec team
Dario Amodei has one employee. The CEO of a $30 billion company directly manages exactly one person.
Anthropic unveils secret 'Mythos' model for AI development in internal testing
Anthropic has a ghost in the machine. While the world watched its legal battles and Pentagon standoffs, the company was quietly building something...
AI Agent Skips Unneeded Tool Call After Observing Zero Precipitation
Most AI assistants are obedient, expensive idiots. They follow the script, run every function you give them, and rack up your API bill.
Design Patterns Use JSON and Shared State to Coordinate Agentic AI
Forget about AI that can hold a conversation. The real work happens when you get multiple AI agents to stop talking and start coordinating.
Industry Shifts to Richer Context for AI Agents, Guided by Human Judgment
Early AI agents ran on a sparse setup: a simple system prompt, a few tool definitions, and hope. That’s changing.
Anthropic adds Slack, Figma, Asana to Claude, touts AI as job substitute
Anthropic just turned Claude into a command center for the tools millions already live inside: Slack, Figma, Asana.
Moonshot's K2 Thinking tops open-source AI, beating GPT-5 and MiniMax-M2
Every open-source AI model claims to be on the verge of beating the big guys. Moonshot's K2 Thinking actually did it, by the numbers.
Anthropic sues Trump as xAI restarts AI coder push, Iran releases AI fakes
Three stories, three distinct battlefields in the AI war, and none of them are quiet.
Anthropic debuts Claude Managed Agents beta for cloud‑hosted AI
Building your own production-ready AI agent has always meant wrestling with a tangle of sandboxing, state management, and tool execution,...
MiniMax's M2.5 costs 1/20 of Claude Opus, covers 30% of HQ tasks
MiniMax claims their M2.5 model costs just one-twentieth of Claude Opus. That alone turns heads.
Claude Code executes DNS‑fetched commands in GitHub repo, evading scans
Modern security scanners check code. They miss what isn't there. Researchers at a German institute just proved how dangerous that gap can be.
NVIDIA OpenShell Controls Execution and Visibility for Autonomous Agents
The promise of autonomous agents is also their biggest problem. You get a tireless digital worker that remembers everything and writes its own code.
Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
Your spreadsheet just whispered to your slide deck. On Monday, Anthropic announced its Claude AI can now maintain shared context between Microsoft...