🚀 Claude Cowork Goes Company-Wide

Anthropic is rolling out its Claude Cowork tool to all paid users on macOS and Windows, making it easier for teams to collaborate with AI. Alongside the expansion, new admin controls like role-based access and usage tracking give companies more control than ever. Let’s dive in!

In today’s insights:

LATEST DEVELOPMENTS

Evolving AI: Anthropic has expanded Claude Cowork to all paid plans on macOS and Windows and added a suite of organizational controls for company-wide deployment.

Claude Cowork started as a way to bring the agentic power of Claude Code to non-developers. Now Anthropic is making it viable at the enterprise level. The update introduces role-based access controls for Enterprise plans so admins can assign specific Claude capabilities to different teams via SCIM or manual setup. Group spend limits let companies set per-team budgets from the admin console. Usage analytics now appear in the admin dashboard and via the Analytics API giving visibility into which teams are adopting Cowork and which workflows are sticking. Expanded OpenTelemetry support means Cowork now emits events for tool calls file access and connector activity compatible with SIEM tools like Splunk. The new Zoom connector brings AI-generated meeting summaries and action items into Cowork so teams can build workflows around their conversations. Admins can also restrict specific connector actions — allowing read access while blocking write operations for example. The desktop app is available now at claude.com/download on both macOS and Windows.

Most companies are already running AI agents across their teams, around 79% of organizations have adopted them in some capacity, However, the governance side has been a real mess. About 67% of executives believe their company has already had a data leak or breach from unapproved AI tools and only one in five companies has anything close to a mature governance model for autonomous agents. That's the gap Anthropic is going after here. Cowork's new controls give IT and finance teams the spend limits, access rules and audit trails they need to actually sign off on a wide rollout rather than quietly letting it spread unchecked. The non-developer angle is also worth watching: the fact that marketing, legal and finance teams are driving most of Cowork's usage suggests that agentic AI is no longer a developer-first story.

Evolving AI: Every week this quarter, something new dropped. A new model. A new workflow. A new way to work faster, think sharper, build better. Keeping up felt like a full-time job.

AI Brief is our live quarterly workshop that cuts through all of it — and turns Q1 2026's biggest AI shifts into tools and workflows you actually use.

This isn't a recap. It's a head start on everyone who just watched Q1 go by and actually start building with AI.

Exclusive coupon code for $50 off for our newsletter readers:
EARLYBIRD-NEWSLETTER

Join AI Brief — April 19th · $199 → $149 (only for our newsletter readers) 

Start Q2 ahead →

Evolving AI: Florida's Attorney General has launched a formal investigation into OpenAI after ChatGPT was allegedly used to plan a deadly campus shooting.

Florida Attorney General James Uthmeier announced Thursday that his office would investigate OpenAI over the alleged role ChatGPT played in the 2025 Florida State University campus shooting. The attack killed two people and injured five. Attorneys for one of the victims claimed last week that the shooter used ChatGPT to plan the assault. Uthmeier posted a statement on X saying his office was "demanding answers" and that subpoenas were forthcoming. OpenAI responded by saying it would cooperate with the investigation and noted that over 900 million people use ChatGPT weekly for legitimate purposes. The FSU case is not isolated. ChatGPT has been linked to a growing number of violent incidents including murders and suicides. A Wall Street Journal investigation found that a man with a history of mental illness regularly used ChatGPT before killing his mother and himself, with the chatbot appearing to reinforce his paranoid thinking. Psychologists have begun referring to this pattern as "AI psychosis."

The FSU shooting is the latest in a pattern that has been building for over a year. ChatGPT has now been linked to suicides, murders and mass shootings, with multiple lawsuits alleging the company knowingly released a version of GPT-4o without proper safety testing despite internal warnings. The bigger legal question being tested here is whether a chatbot is a passive tool, like a search engine, or something closer to an active participant in a user's decisions. That distinction could redefine liability standards across the entire tech industry. If courts land on the latter, it forces every AI company to rethink not just their guardrails but their entire product design. And given that OpenAI rolled out parental controls only after facing a Senate hearing and public pressure, the question of whether the industry will self-correct without legal force pretty much answers itself.

2026 will redefine IT as a strategic driver of global growth. Automation, AI-driven support, unified platforms, and zero-trust security are becoming standard, especially for distributed teams. This toolkit helps IT and HR leaders assess readiness, define goals, and build a scalable, audit-ready IT strategy for the year ahead. Learn what’s changing and how to prepare.

Download the Toolkit

Evolving AI: OpenAI launched a new $100/month plan built around its Codex coding tool, a direct challenge to Anthropic's Claude Code.

OpenAI now offers a $100/month plan sitting between its $20 Plus tier and $200 Pro tier. The new plan is designed for developers who rely heavily on Codex but found Plus too limited and the $200 plan too expensive. The $100 tier offers 5x the Codex capacity of Plus. The $200 plan still exists and provides 20x Plus limits for the most demanding workflows. Both higher-tier plans are ad-free. The free and $8/month Go plans still carry ads.

There is one catch: OpenAI is running elevated Codex limits on the $100 plan only through May 31. After that date limits will likely tighten. Codex already has strong momentum, over 3 million users per week with usage up 70% month over month.

The AI coding tools race has gotten very real, very fast. Codex went from handling roughly 5% of Claude Code's weekly usage in September 2025 to about 40% by January 2026, and OpenAI has been openly redirecting resources toward it because of pressure from Anthropic. The $100 price point isn't a coincidence either, Anthropic has had that tier for a while, and OpenAI is now planting a flag right next to it. What's interesting is that both tools are converging in capability, so the fight is shifting away from "which model is smarter" toward who offers better value at each price tier. With the AI coding market projected to hit $91 billion by 2035, the subscription structure OpenAI builds now could shape how millions of developers choose their daily tools for years to come.

With OpenAI lowering its subscription fee for Codex, this demo shows why that’s a big deal. Instead of just assisting, Codex runs tasks in parallel workspaces, letting you build features while it codes alongside you. It’s a shift from AI as a tool to AI as a teammate, and that’s where dev workflows and the future are heading.

QUICK HITS

👨‍⚖️ Ohio man becomes first to be convicted under new AI statute for sexually explicit images.

🇬🇧 OpenAI pauses UK data centre deal over energy costs and regulation.

🌐 The new, AI-powered Google Finance is expanding to more than 100 countries.

🚀 Claude’s new “Advisor Strategy”.

 *partner link