Claude

Claude Duet is a new app where Claude messages you first, and only when something needs you. Each morning it sends a short brief on your top priorities. Hand it a bigger job by voice note, put your phone down, and it keeps working. Before it sends, pays, or signs anything, it waits for your okay. It’s an early research preview and we’re still tuning it based on user feedback. What do you want to know about how Claude Duet works? Join the waitlist on Pro and Max plans in the US: claude.com/claude-duet

In my conversation with Daniela Amodei, president of Anthropic, we discussed one of the company’s core cultural values: holding light and shade, focusing on the incredible potential benefits AI can bring about but also discussing and mitigating the real risks. This value of holding light and shade is not just true about AI, it's true about all of us. We’re all a mixture of better angels and lesser angels. Daniela and I discussed why this moment isn’t just about how AI can constantly improve, but also what we can do to nurture the best version of ourselves. Watch our conversation at the link in the comments.

Claude Models on Google Cloud (Anthropic managed) is now generally available. Google Cloud customers get the full set of Claude Platform features, with Google Cloud authentication, billing, and commitment retirement. Anthropic operates the service, with new features and betas shipping the same day they go live on the native Claude Platform. Build and deploy agents with Claude Managed Agents (beta), or use features like fast mode, code execution, skills, the MCP connector (beta), and more. Claude also remains available through the Google managed offering, where Google Cloud is the data processor. Teams with strict regional data residency requirements or that want native Google Cloud integrations like Model Armor should consider Claude Models on Google Cloud (Google managed). Read more: https://lnkd.in/ehFB7w4a

Starting next week, sessions across Claude web, desktop, and mobile apps will run in auto mode by default on Pro, Max, and Team plans. Rather than asking you to approve each step, an automated check blocks risky actions. Switch back to manual anytime.

Whatever you’re curious about, there’s an artifact you could make for it with Claude. People at Anthropic have been making them for their own curiosities, like an app that tracks bird migration on the west coast, or one that turns handwriting into a downloadable typeface.

We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, and keeps working when you step away. We designed it around Claude Opus 5.5, which gives it the judgment to decide what needs you, what can wait, and when to check with you first. Before it sends something sensitive, makes a payment, or deletes anything, it shows you exactly what it’s about to do and waits for your approval. It carries over the memory, skills, and connected apps from your Claude account, so setting it up takes minimal work. We built it for professionals and small business owners with more on their plates than they can track. Claude Duet draws on years of our safety and security work on agents. It never directly sees your passwords, a new classifier guards against sharing sensitive information, and the important decisions stay with you. Claude Duet is an early research preview for Pro and Max subscribers in the US, on iOS, Android, and web. It shares your plan’s usage limits, and it might still misread a thread or misjudge what’s urgent. We’ll keep tuning it based on what early users tell us. Join the waitlist: https://lnkd.in/eqcK4WXZ Read more: https://lnkd.in/eUchXTFy

Claude Security now keeps up with your code as it changes. Connect your repositories once. Claude runs a deep first scan, then scans continuously, so your list of findings stays close to the current state of your code. A project can hold several repositories across any git provider such as GitHub and GitLab, which lets Claude catch issues that live between services, like a backend that trusts input a client was supposed to validate.  Most fixes happen outside the tool that found the issue. Claude now checks open findings against the code that changed and marks them fixed when the fix lands, wherever your team made it. If the issue shows up again, the finding reopens. Your team can also tell Claude what a scanner can’t see from the code alone (how it’s deployed, which risks you’ve already accepted) and edit the threat model Claude writes for each project. Findings reach your ticketing and reporting tools through a new API, and through an MCP connector in Claude and Claude Code. Claude Security is in public beta on Claude Enterprise, and eligible organizations receive up to $25,000 in free usage credits per month, for 2 months, through December 31, 2026. Terms apply. Read more: https://lnkd.in/eEmy-kn9

When Luis and Laura started SoFlo Solar, Laura had never written code. She built the first version of the company’s operating platform with Claude in eleven days. Today, their Miami-based company installs new solar systems and repairs the ones other companies left behind. Read more stories: claude.com/customers

Doug is a product engineer at Anthropic. He writes his launch reviews himself, then asks Claude to pull the numbers, chart them in the doc, and check whether he got it right. That’s the idea behind Claude Docs: Claude works in the doc alongside your team, down to answering questions in the comments. Here's how Doug uses it for standups and launch reviews.

Jonothan Hunt loves old cameras for all of their flaws, so he spent a weekend with Claude Code building one with flaws of his own design. It's slow, it's grainy, and every photo is a small surprise.