Figma Uses AI to Transform Digital Design and Collaboration

TL;DR

• AI is embedded across Figma’s design workflows, accelerating prototyping and collaboration.
• Figma Make enables prompt-to-app production-grade code generation; Dev Mode MCP Server translates mocks into production-ready code with full context.
• Figma uses OpenAI APIs to power FigJam AI and image generation; ChatGPT Enterprise is deployed company-wide to boost AI fluency.
• The product design workflow now supports cross-modality, language, visuals, and code, keeping human craft central.
• A real-world internal example shows non-technical employees building usable tools with AI, lowering the barrier to experimentation.

Context and background

In August 2025, OpenAI’s Executive Function series featured a conversation with David Kossnick, Head of AI Products at Figma, to explore how AI is transforming design and how Figma is positioning itself in that shift. The interview emphasizes that AI is not just about pixels; it is a platform shift and a core capability that rethinks workflows from first principles. Figma’s vision centers on making design faster, more intuitive, and accessible to more people by embedding AI across products—from in-product text editing and image generation to auto-renaming layers and site visuals. The discussion also highlights Figma Make, a prompt-to-app tool that can generate production-grade code from language, images, or structured frames, enabling designers, PMs, engineers, and marketers to prototype and express ideas beyond traditional technical barriers. OpenAI interview. Figma frames AI as both a tool and a platform shift. The company emphasizes that collaboration is at the core of its design philosophy, with ideas and code flowing in a multiplayer environment that supports real-time co-creation even when AI is involved. The interview underscores that AI should augment human judgment, empathy, and taste—qualities that make designers the true pilots of the process. The integration of AI into end-to-end workflows, including Figma Make and the Dev Mode MCP Server, is presented as a systematic approach to embedding AI across design, development, and delivery. OpenAI interview. Internal adoption and fluency are positioned as strategic priorities. Figma has pursued dogfooding, company-wide experiments, and structured programs to help employees gain hands-on experience with AI tools. Initiatives like the Great Figma Bake Off, Maker Weeks, and the rollout of ChatGPT Enterprise are cited as core mechanisms to reduce barriers to experimentation while ensuring data guardrails and secure usage. This interview also touches on the evolution of designer and developer collaboration in an AI-augmented workflow—from handoff artifacts to fully integrated AI-assisted coding experiences. Dev Mode and MCP are described as ways to translate mockups into production-ready code with minimal friction, reinforcing Figma’s commitment to making collaboration between disciplines more seamless and effective. For more context and direct quotes from the discussion, see the OpenAI interview linked above. OpenAI interview.

What’s new

• Figma Make: A prompt-to-app tool that generates production-grade code from language, images, or structured frames, enabling coders and non-coders alike to prototype and express ideas beyond traditional barriers.
• Dev Mode MCP Server: A system that supports a tighter handoff by using structured data (e.g., CSS tokens) and enabling a coding agent to translate mocks into production-ready code with full context—eliminating manual copy-pasting.
• MCP (Code Translation Agent): Extends Dev Mode by letting developers invoke a coding agent that translates mocks into production-ready code with full project context.
• Cross-modality workflows: AI-powered support for language, visuals, and code to let teams work across disciplines without sacrificing their primary strengths.
• Image generation in FigJam and Slides: Teams can co-create brand-aligned visuals and iterate side-by-side.
• AI-enabled collaboration: Real-time multiplayer AI experiences where multiple users co-create with an AI assistant in the same file and avatars are visible.
• AI fluency and governance: Company-wide adoption of ChatGPT Enterprise, Maker Weeks, and structured experimentation with guardrails to enable safe exploration.
• Internal use cases: A HR team member built a Workday API-based game using Figma Make in two hours, illustrating how AI lowers the barrier to deploying usable internal tools.

Why it matters (impact for developers/enterprises)

• Accelerated idea-to-prototype loops: By embedding AI across design and development, teams can test more ideas faster and validate value with less friction. The ability to prototype with AI-generated code helps bridge the gap between concept and execution.
• Enhanced collaboration: Figma emphasizes multiplayer experiences, ensuring AI augments rather than replaces human collaboration. Real-time co-creation with AI supports diverse skill sets and cross-disciplinary teamwork.
• Broader designcraft and inclusion: AI is used to expand craft to language, visuals, and code, enabling more people to participate in product creation without requiring deep coding expertise.
• Lowered barriers to experimentation: Initiatives like Great Figma Bake Off and Maker Weeks demonstrate a culture where employees can safely explore new AI workflows, accelerating the discovery of real value.
• Real-world velocity: Early internal successes, such as a non-technical employee building a Workday-based game, show how AI workflows can unlock practical, deployable tools outside traditional engineering teams.

Technical details or Implementation

Figma’s AI strategy is anchored in a set of integrated capabilities that span design, code, and data handoff. Notable components include:

• Figma Make: A prompt-to-app tool that enables production-grade code generation from language, images, or structured frames. This capability supports designers, PMs, engineers, and marketers in prototyping and expressing ideas without being blocked by technical barriers.
• Dev Mode and MCP Server: Dev Mode streamlines handoff with structured data such as CSS and tokens; MCP advances this by enabling developers to invoke a coding agent that translates mocks into production-ready code with full context, removing the need for manual copy-pasting.
• Cross-modality editing: Figma supports editing across language, visual, and code layers, allowing users to tailor AI-generated results to match their creative vision.
• AI agents and code layers: The platform introduces AI agents capable of handling repetitive tasks, while still allowing full editing of language, visuals, and code to preserve craft and intent.
• OpenAI APIs and FigJam AI: OpenAI APIs power FigJam AI and image generation on the platform, underscoring a tight integration between AI capabilities and Figma’s collaborative design tools. Figma has also deployed ChatGPT Enterprise across the organization to foster AI fluency.
• Multiplayer AI experiences: Figma emphasizes real-time collaboration with AI, where two people can work in the same file, see avatars, and co-create with an AI assistant.
• Internal programs for fluency: Initiatives like the Great Figma Bake Off and Maker Weeks promote hands-on experimentation, while guardrails provide a compliance fast path for safe AI exploration. Table: key AI-enabled tools in Figma (illustrative overview) | Tool | What it does | User impact |---|---|---| | Figma Make | Prompt-to-app tool generating production-grade code | Expands prototyping reach to non-developers and speeds iteration |Dev Mode | Structured data handoff with tokens/CSS; supports code translation | Improves handoff quality and reduces manual effort |MCP Server | Coding agent translating mocks to production-ready code with context | Enables engineers to generate deployable code from designs |OpenAI APIs powering FigJam AI | AI-powered features and image generation in FigJam | Enhances collaboration and visual ideation | Key takeaways
• AI is viewed as a platform shift integrated into Figma’s core product, not just a set of features.
• The design-to-build workflow is being reimagined to support cross-modality collaboration across language, visuals, and code.
• Real-world, non-technical use cases demonstrate that AI can enable deployable tools with minimal engineering.
• Internal programs and guardrails help scale AI fluency while maintaining security and data use standards.
• The future of design at Figma emphasizes craft, human judgment, and the ability to explore and execute ideas end-to-end with AI support.

References

• OpenAI interview