Gottheimer Presses Anthropic on Repeated Code Leaks
Rep. Josh Gottheimer pressed Anthropic's CEO on repeated Claude Code leaks, citing national security risks. Anthropic called it a packaging error, not a breach, with no customer data exposed.
Anthropic
Staying ahead of the AI race with a look at Anthropic’s mission for safe AI. We break down the technical shifts in Claude and the future of LLMs.
Anthropic Re-Leaks Claude Code Source via 60 MB npm Map File
Anthropic’s Claude Code npm package contained a 60 MB source-map file that let anyone reconstruct its full TypeScript source. It’s the second time the company has made the same packaging mistake.
Anthropic Leaks 'Claude Mythos': Cyber-Supermodel Needs Babysitters
A mis-configured blog tool spilled Anthropic’s next-gen model, Claude Mythos, plus warnings it could outrun cyber defenders. The company calls it a step-change in capability—and a cybersecurity headache.
AI Unearths 22 Firefox Flaws in Two Weeks, But Exploits Remain Costly to Craft
Anthropic's AI identified 22 Firefox vulnerabilities in two weeks, but exploit generation remained limited despite $4,000 in API costs. Mozilla validated AI-assisted security as a new tool.
From 'Dr. Google' to AI Chatbots: The $230 Million Weekly Gamble on Health Data
AI chatbots handle 230 million weekly health queries, but face challenges in translating medical jargon and securing patient data.
The Claude C Compiler: How AI Just Walked Into the Engineering Room
Anthropic's Claude C Compiler didn't just write code — it architected an entire system. That's a genuine milestone. But the deeper story is in what it got wrong, what it couldn't invent, and what it means for every engineering team navigating the shift from writing software to directing it.
AI Cold War 2.0: Can U.S. Tech Corps Outspend China’s Cheap AI in the Global South?
The U.S. Tech Corps aims to counter China’s AI dominance in the Global South, but cost disparities and infrastructure gaps threaten its viability.