New Anthropic research: How are people using AI systems in the real world?
We present a new system, Clio, that automatically identifies trends in Claude usage across the world.
New Anthropic research: Alignment faking in large language models.
In a series of experiments with Redwood Research, we found that Claude often pretends to have different views during training, while actually maintaining its original preferences.
Results of our jailbreaking challenge:
After 5 days, >300,000 messages, and est. 3,700 collective hours our system got broken. In the end 4 users passed all levels, 1 found a universal jailbreak. We’re paying $55k in total to the winners.
Thanks to everyone who participated!
Claude 3.7 Sonnet is a state-of-the-art model for both coding and agentic tool use.
In developing it, we’ve optimized somewhat less for math and computer science competition problems, and instead shifted focus towards real-world tasks that better reflect the needs of our users.
Proud to share what my team and I have been working on! I use Claude Code almost every day, and wouldn't want to code without it. Just getting started.
Three more features for Claude Code today:
First, auto-accept mode.
Press Shift+Tab to toggle on auto-accept for file edits. This reduces friction while keeping you in the driver’s seat—you can tab in/out of auto-accept mode anytime.
A new feature release for Claude Code: you can now quickly add memories to the CLAUDE.md file using the # key.
Tell Claude Code to remember specific tips and best practices, and it’ll reference that information as context in future sessions.
Here’s how it works:
Introducing a new Max plan for Claude. It’s flexible, with options for 5x or 20x more usage compared to our Pro plan.
Plus, priority access to our latest features and models:
I've been using our internal prototypes of our MCP-enabled research workflow. It's amazing what Claude digs up. No more manual context stuffing! Hook up your data and give it a shot 🪏🪏🪏
I've been using versions of Opus 4 for coding for a bit now. The numbers in the chart don't do it justice. Absolute game changer in staying on track, writing maintainable code, and generally doing what I want and expect rather than reward hacking. A+ job, team!
At a time when people are understandably focused on the daily chaos in Washington, these articles describe the rapidly accelerating impact that AI is going to have on jobs, the economy, and how we live. axios.com/2025/05/28/ai-…
Since we originally built Claude Code as an internal tool, we've heard a ton of questions about how our teams use it at Anthropic.
Here’s an inside look on how our teams—from product engineering, to growth marketing, to legal—use Claude Code:
We're hiring for the Claude Code team! In particular, we're looking for a Systems Engineer (job-boards.greenhouse.io/anthropic/jobs…) and an Eng Manager (job-boards.greenhouse.io/anthropic/jobs…). The Claude Code team is having a blast, come join us :)
Local MCP servers can now be installed with one click on Claude Desktop.
Desktop Extensions (.dxt files) package your server, handle dependencies, and provide secure configuration.
Love to see this!! Since we first prototyped Claude (and before it was named that), one of the most surprising and impressive things was how empathetic it seemed to be, and that's only improved over time. Try it and see ❤️