Anthropic is teaming up with PayPal for an upcoming webinar that’s aimed squarely at one of the hardest problems in modern IT: how do you roll out an AI desktop agent like Claude Cowork across an entire enterprise without losing control of cost, security, or governance?

The session, scheduled for April 16, 2026, at 10:00 am PT, is designed for the people who actually own these rollouts inside large companies: CIO organizations, IT admins, heads of engineering, and the operations leaders who get asked to “make AI happen” for thousands of employees at once. Rather than focusing on abstract AI hype, Anthropic and PayPal are positioning this as a practical, live walkthrough of what it looks like to turn Cowork from a pilot project into something that becomes part of everyday work for entire teams.

At the center of all this is Claude Cowork, Anthropic’s new agentic AI for knowledge work. Cowork runs on the desktop, connects to local files, folders, and the apps people already use, then autonomously executes multi‑step tasks: cleaning up chaotic file structures, consolidating data into spreadsheets, drafting documents and presentations, and more. Instead of forcing employees to break work into dozens of prompts, Cowork is designed around outcomes: you describe the goal, point it at the right sources, and it figures out the steps. That agentic model has already generated a lot of excitement among early adopters, but the real challenge now is scale—moving from a handful of power users to whole departments and, eventually, the full enterprise.

That’s the gap this webinar is trying to address. Anthropic is now making Cowork generally available across paid plans on macOS and Windows, backed by the kinds of controls big organizations demand before they let an AI agent anywhere near real production data. The pitch is simple: one deployment of the Claude desktop app can cover the entire company, from Cowork for non‑technical knowledge workers to Claude Code for engineering teams, all under a unified admin and compliance model.

To make that possible, Anthropic has been rapidly layering in enterprise‑grade guardrails. The webinar will showcase new role‑based access controls, allowing admins to create groups, assign custom roles, and tightly scope which capabilities each team can use—whether that’s Cowork, Code, web search, external connectors, or other tools inside the Claude ecosystem. That means a finance team can get deep access to document handling and spreadsheet workflows, while a legal team might have stricter connector permissions and a more conservative configuration around external data access.

Closely tied to that is the question of cost. Anthropic is explicitly addressing the budget side with group‑level spend caps, giving enterprises a way to manage usage by team without throttling individual productivity. Rather than treating AI as an uncontrolled expense, these per‑team budgets make it easier for IT and finance leaders to forecast spend, run internal chargeback models, and justify expansions based on real usage and ROI.

Monitoring is another big theme. Anthropic is extending usage analytics through an admin dashboard and an Analytics API, plus OpenTelemetry support, so organizations can plug Cowork’s activity into their existing observability stacks. For large enterprises, that’s not a nice‑to‑have—it’s how they track adoption, spot misuse, and measure whether AI agents are actually saving time or just becoming another shiny object. Granular controls over connectors and tools further reduce risk, letting administrators restrict which business systems Cowork can touch and how far its automations can go.

What makes this webinar particularly interesting is the presence of PayPal as a reference customer. The speaker lineup includes Jarred Keneally, Senior Director of Product Management, AI Technology at PayPal, alongside John Lopus from Anthropic’s Applied AI team and Amber Yin from Customer Success. Rather than a purely vendor‑driven demo, attendees will hear directly how a major global payments company is approaching a Cowork rollout. PayPal plans to share what’s working, what they’re monitoring, and how Anthropic’s admin features have shaped their deployment plan.

The Anthropic–PayPal collaboration isn’t coming out of nowhere. PayPal has already partnered with Anthropic to integrate its payment services with Claude, allowing businesses to manage transactions, subscriptions, disputes, and financial insights via conversational commands. Using Anthropic’s Model Context Protocol and a directory of connected tools, Claude can tap into real‑time PayPal data to generate financial reports, review disputes, and even help with invoicing and sales analysis. For enterprises, that kind of integration hints at where Cowork is headed: an AI agent that doesn’t just summarize information, but acts across core systems—finance, operations, and customer support included.

From an enterprise buyer’s perspective, this webinar looks less like a generic product pitch and more like a deployment playbook. Anthropic is promising a phased rollout model that has already worked for other large teams: start with targeted pilots, expand to specific departments, then broaden access as governance, training, and telemetry mature. Combined with features like org‑wide branding inside Cowork, tailored home experiences, and cross‑platform support, the message is that Claude is ready to move from isolated experiments to a consistent, company‑wide AI layer.

For IT leaders who have been cautious about agentic tools running on employee desktops, the combination of least‑privilege access, per‑team budget limits, and detailed analytics is a strong signal that Anthropic understands the realities of enterprise governance. And with PayPal on stage talking candidly about what they’re watching during rollout—adoption, productivity gains, and risk—the session is likely to resonate with organizations that have been waiting for a credible blueprint before going beyond small‑scale trials.

Registration is open via Anthropic’s site, with standard fields for name, business email, job title, and company, and the company notes that attendance is covered by its usual privacy policy. For enterprises that are serious about AI agents but wary of losing control, this webinar offers a timely look at how one of the biggest names in payments is tackling the same questions—and how Anthropic is adapting Claude Cowork to meet those expectations.