Anthropic has released a significant update for its Claude Code product, introducing a feature called Artifacts that transforms the output of an AI coding session into a live, interactive, and shareable web page. Available to users on the Claude Team and Enterprise subscription plans, this new capability allows engineers to generate dynamic HTML pages—such as dashboards, app prototypes, or data visualizations—directly from their terminal or desktop app. The resulting page is hosted at a unique URL that can be shared with colleagues, who can then watch it update in real time as Claude Code continues to work autonomously or under the user’s direction, pulling in data from connected codebases and monitoring tools.
How Claude Code Artifacts Work: From Terminal to Live Canvas
At its core, the Artifacts feature acts as a dynamic translation layer between a developer’s command-line work and the stakeholders who need to understand it. Instead of manually wiring up external data sources or standing up temporary infrastructure, an engineer can prompt Claude Code to investigate a problem—such as a user drop-off after a recent release—and the agent will execute the necessary queries, build an interactive funnel dashboard, and present it as a live web page. As the AI refactors code or updates its analysis, the open webpage refreshes in-place at the same URL, with every change recorded in a version history that teammates can roll back or review on desktop or mobile.
This approach effectively replaces static status reports, whiteboard diagrams, and manual bug walkthroughs with a secure, self-updating visual tool. The output is not a static export; it is a living document of the agent’s reasoning and progress.
Anthropic’s Claude Code Artifacts vs. OpenAI’s Codex Sites: A Strategic Divide
This update arrives just over two weeks after OpenAI introduced a similar enterprise hosting feature for its Codex platform called “Sites,” setting up a direct competitive dynamic between the two leading AI companies. While both features aim to make AI-generated work more accessible to non-technical teams, their underlying philosophies and technical architectures diverge sharply.
OpenAI’s Sites is designed as a platform-as-a-service, generating durable, full-stack web applications. According to its documentation, Codex Sites outputs projects as Cloudflare Worker-compatible ES modules and supports persistent backend infrastructure, including D1 relational databases and R2 object storage. A Site can handle public sign-ins, integrate with external identity providers, and support granular access controls for specific workspace groups. It uses a two-stage publishing process tied to Git commits, making it a production environment intended to replace internal SaaS tools.
Anthropic’s Claude Code Artifacts, by contrast, deliberately avoids the backend. The documentation is explicit: “An artifact is a capture of work, not an application.” Each artifact is a single, self-contained HTML page capped at 16 MiB. To guarantee organizational security, Claude wraps the file in a strict Content Security Policy (CSP) that blocks all external network requests—no external scripts, fonts, stylesheets, or API calls are permitted. All CSS and JavaScript must be inlined, and images must be embedded as data URIs. Artifacts cannot store form input, call an API at view time, or serve multiple routes.
This technical limitation is a deliberate philosophical choice. While OpenAI aims to spin up persistent software portals for the entire company, Anthropic is keeping Claude Code anchored in ephemeral, highly secure technical workflows. Claude Artifacts are not meant to be software; they are meant to replace the friction of communicating technical work with secure, self-updating visual tools that never leak live data outside the corporate boundary.
Enterprise Security and Licensing: Keeping the Codebase Private
Because these agents sit at the nexus of proprietary company data and live codebases, licensing and access controls are a primary concern for enterprise buyers. Both Anthropic and OpenAI have opted for closed, proprietary licensing models for these new visual workspaces. Unlike permissive open-source software (such as MIT or Apache 2.0) or strict copyleft licenses (like GPL), neither Claude Code Artifacts nor Codex Sites can be independently forked or hosted. Enterprise clients do not maintain code-level ownership over the rendering engine or integration nodes; both operate strictly within their respective creators’ managed infrastructures.
To make this vendor-managed approach palatable to compliance teams, both companies have heavily prioritized organizational security. Anthropic ensures every artifact is private to its author by default and cannot be made public to the broader internet. When an engineer shares a link, it is viewable exclusively by authenticated members of their specific organization. System administrators retain ultimate authority, managing access through org-level toggles, role-based scoping, and explicit retention policies, with oversight available through a centralized compliance API.
OpenAI takes a similarly gated approach, rolling out Sites primarily for ChatGPT Business and Enterprise workspaces. System administrators must explicitly enable Sites via role-based access control (RBAC) for Enterprise tiers. However, because Codex Sites functions more like a hosted web application, its access controls are slightly more granular. When an engineer prepares to share a deployed URL, they can apply specific access modes: restricting the site to just themselves and workspace admins, opening it to all active users in the workspace, or limiting access to custom user groups. OpenAI also provides a dedicated Sites panel to manage runtime environment variables and secrets securely, ensuring those keys do not have to be committed to local source files.
What This Means for Developers and Enterprise Teams
The introduction of visual, self-updating UI layers to command-line agents is fundamentally altering how developers view their own workflows. As AI handles the raw syntax and automates the reporting, the friction of communicating technical work to stakeholders is vanishing. Boris Cherny, the Lead and creator of Claude Code, highlighted the utility of the update in a post on X, noting that he has been using Artifacts for everything from visual explanations of tricky code and system diagrams to data analyses and dashboards shared with his team.
This sentiment is practically demonstrated in Anthropic’s launch materials. In one scenario, an engineer prompts Claude Code to investigate user drop-offs since a previous software release. In seconds, the agent executes an SQL read, builds an interactive drop-off funnel dashboard, and diagnoses that “Pro accounts stall at the export sheet.” The AI then proposes UI fixes, updates the live charts as the code is refactored, and generates a secure link that a manager can instantly open via mobile.
By turning the terminal into a live, collaborative canvas, Anthropic is proving that the most valuable output of an AI coding assistant is not just the code itself—it is the context, the reasoning, and the ability to share that work instantly.
Who Should Try This Now
For engineering teams already using Claude Code on a Team or Enterprise plan, Artifacts is available immediately and requires no additional configuration. The feature is best suited for developers who need to communicate complex technical findings to non-technical stakeholders without building separate dashboards or scheduling walkthroughs. It is not designed for teams looking to deploy production-grade internal tools—that remains the domain of OpenAI’s Codex Sites. For those who need a secure, ephemeral, and highly visual way to share the output of an AI-driven coding session, Claude Code Artifacts represents a practical and immediately useful step forward.
