Anthropic Launches Claude Design: AI Tool That Could Disrupt Adobe, Figma & the Future of Creative Work

Anthropic has officially introduced Claude Design, a new experimental product that expands its AI assistant beyond text and coding into visual content creation. Announced on April 17, 2026, the feature is currently available in research preview and allows users to create polished presentations, prototypes, slides, one-pagers, product mockups and marketing materials using simple natural language prompts.

The launch marks a major step in Anthropic’s broader strategy to position Claude as a complete workplace AI assistant rather than only a chatbot or coding tool. Claude Design is powered by the company’s most advanced vision model, Claude Opus 4.7 and is being rolled out to Claude Pro, Max, Team and Enterprise subscribers at no additional cost. Anthropic says access will be gradually expanded throughout the day for eligible users.

According to Anthropic, Claude Design is built to help both professional designers and non-designers. Designers often have limited time to test multiple creative directions, while founders, marketers and product managers may struggle to turn ideas into polished visual output without design expertise. Claude Design aims to solve both problems by letting users simply describe what they need and allowing Claude to generate a first version that can then be refined through conversation, inline comments, direct edits, or custom sliders.

One of the standout features is automatic brand learning. During onboarding, Claude can read a company’s codebase and design files to build a design system that understands brand colors, typography and components. This allows future projects to automatically match a company’s visual identity without requiring manual setup. Anthropic says teams can also maintain multiple design systems and improve them over time.

Users can begin projects in several ways. They can start with a text prompt, upload files such as DOCX, PPTX and XLSX documents, import images, or even point Claude directly to their codebase. A web capture tool also allows users to pull elements from websites so prototypes closely resemble real products.

The company says Claude Design is already being used for realistic prototypes, product wireframes, design explorations, investor pitch decks, marketing collateral and even frontier design projects involving voice, video, shaders, 3D and built-in AI. Once a design is complete, users can export it as PDF, PPTX, standalone HTML, or directly to Canva. Designs can also be handed off to Claude Code for implementation through a single bundled instruction.

Anthropic says collaboration is built directly into the system. Teams can keep documents private, share them within an organisation using internal links, or allow edit access so colleagues can modify designs and chat with Claude together inside the same project.

The announcement has also triggered immediate attention across the design software industry. Reports suggest investor concern around how Claude Design could affect traditional platforms such as Adobe and Figma, particularly because the feature is bundled inside existing Claude subscriptions rather than sold as a separate design tool. Its ease of use and automation-first workflow have positioned it as a potential challenger to conventional design software.

For Enterprise customers, Claude Design is disabled by default and administrators must manually enable it through organisation settings. Anthropic has also indicated that more integrations with external tools will be introduced in the coming weeks.

While still in its early research preview stage, Claude Design signals a much larger shift in the AI industry. Companies are no longer limiting AI assistants to writing and coding tasks. Instead, they are racing to combine design, productivity, collaboration and execution into unified platforms. With Claude Design, Anthropic is making it clear that it wants Claude to become the central operating system for modern work.