Anthropic just turned Claude into an enterprise app store. The company shipped 12 new connectors for Cowork. Google Workspace (Calendar, Drive, Gmail), DocuSign, Apollo and more, plus a feature that lets admins build private plugin marketplaces. Think of it as an internal app store for AI agents, where IT controls which plugins employees can access and auto-installs tools for specific roles. They also added 10 new plugin templates covering HR, design, engineering, investment banking, and wealth management. The amazing part is Claude can now pass context seamlessly between Excel and PowerPoint without you restarting the conversation.

Today in AI:
→ Claude Code now works from your phone, tablet, or any browser
→ Gemini can order your Uber and DoorDash automatically
→ How to continue Claude Code sessions from anywhere

Let's dive in..

What if you could start a coding task at your desk and finish it from the couch?

Anthropic just shipped Remote Control, and it solves a problem I needed. You can now continue a local Claude Code session from your phone, tablet, or any browser without anything moving to the cloud. Your filesystem, MCP servers, tools, and project configuration all stay on your machine. The web and mobile interfaces are just a window into that local session.

Here's how it works:

The conversation stays in sync across all connected devices. You can send messages from your terminal, browser, and phone interchangeably. If your laptop sleeps or your network drops, the session reconnects automatically when your machine comes back online.

I keep thinking about the use cases here. You're deep in a refactor, dinner's ready, and you want to keep an eye on what Claude's doing. Or you're on the train and want to review what your agent accomplished while you were in meetings. The session doesn't care where you are, it just keeps running.

One limitation worth knowing: your laptop needs to stay awake and online. Network outages longer than about 10 minutes will timeout the session. But for the "I need to step away but don't want to lose momentum" scenario, this is exactly right.

For your team: If your developers are already using Claude Code, Remote Control is free with Pro and Max plans. Worth testing for anyone who context-switches between desk and mobile throughout the day.

Why does business news feel like it’s written for people who already get it?

Morning Brew changes that.

It’s a free newsletter that breaks down what’s going on in business, finance, and tech — clearly, quickly, and with enough personality to keep things interesting. The result? You don’t just skim headlines. You actually understand what’s going on.

Try it yourself and join over 4 million professionals reading daily.

Check it out

Google just pushed Android deeper into the agent era.

Starting in March, Gemini can handle multi-step tasks like ordering an Uber or reordering your last DoorDash meal without you tapping through the apps yourself. Long-press the power button, tell Gemini what you want, and watch it work. The AI scrolls, taps, and types in a secure sandbox that can't access the rest of your device.

What makes this different from previous assistant features:

I'm genuinely curious how this plays out. The promise of phone automation has been around forever, but it always felt clunky like macros that broke whenever an app updated. Gemini's approach of running in a sandboxed window with live monitoring feels more trustworthy. You're not handing over the keys; you're watching a very fast assistant do the tedious parts.

For your team: If you're building mobile apps, expect users to start asking "why can't I just tell my phone to do this?" The bar for multi-step workflows is about to change.

Remote Control lets you pick up Claude Code sessions from anywhere without losing your local environment. Here's how to set it up:

Tips for best results: Keep your laptop plugged in if you'll be away for a while; sleep mode pauses the session until it wakes

🔓 Hacker used Claude to steal 150GB of Mexican government data. An attacker jailbroke Claude by framing requests as a "bug bounty" program, convincing the AI to act as an "elite hacker." The result was tax records for 195 million people, voter registration data, and government employee credentials. Anthropic banned the accounts and says Opus 4.6 now includes probes to disrupt similar misuse.

🎙️ ElevenLabs raises $500M at $11B valuation. Series D led by Sequoia. Voice cloning went from novelty to infrastructure fast. ElevenLabs now powers podcasts, audiobooks, and AI agents across thousands of apps.

⚠️ Meta's AI safety team warned about Llama 4 risks—company released it anyway. Internal documents show safety researchers flagged concerns that the model hadn't been sufficiently tested before release. Commercial pressure won. The pattern feels familiar: ship fast, patch later. Whether that's sustainable for frontier models is an open question.

For a full list of 1500+ AI tools, visit our Supertool Directory