Figma is moving to formalize how artificial intelligence agents interact with its platform, introducing new features that appear designed to give teams more control over automated actions while broadening the company's role in the design-to-code pipeline. The announcement comes as the collaborative design software maker also opens its code layers feature to beta users, underscoring a wider product strategy built around AI-enabled creation rather than passive assistance.
Agent Controls
The new agent-focused features are notable not simply because they add automation, but because they attempt to define the operating rules for that automation. In the current wave of frontier AI products, companies are racing to embed agents into everyday workflows, yet many enterprise users remain wary of systems that can act too freely inside mission-critical environments. Figma's approach suggests an effort to balance speed with governance, allowing agents to be useful without becoming opaque or unpredictable.
For a platform used by designers, product teams, and developers working in shared environments, that distinction matters. Figma's value has long rested on collaboration, versioning, and the ability to keep creative work visible across teams. Introducing agent behavior into that setting raises practical questions: What can an agent edit? When should it ask for approval? How much context should it retain? By framing the features around how agents work, Figma is effectively positioning itself as a platform architect for AI behavior, not just a beneficiary of third-party models.
That strategy also reflects a broader shift in the software industry. As generative AI moves from novelty to workflow infrastructure, the competitive edge is increasingly about control surfaces, permissions, and integration depth. Companies that can define the boundaries of agentic action may be better placed to win enterprise trust than those that merely bolt on chat interfaces or generic copilots.
Code Layers Beta
Alongside the agent announcement, Figma is releasing its code layers feature in open beta, a move that could further tighten the link between design intent and production-ready output. Code layers are part of a larger industry push to reduce the friction between what designers create and what engineers ultimately ship. In practical terms, the feature points toward a more unified environment where design elements can carry structured code logic rather than existing only as static visual assets.
The open beta release is significant because it invites broader testing at a moment when the market is still deciding how far AI-assisted development can go in real production settings. For Figma, beta access is both a product milestone and a feedback mechanism. It allows the company to observe how teams use the feature, where it breaks down, and how much trust users are willing to place in AI-generated or AI-managed code structures.
The timing is also strategic. Design software is no longer competing only with other design tools. It is increasingly competing with code editors, prototyping platforms, and AI-native development environments that promise to compress the path from idea to application. By advancing code layers while defining agent behavior, Figma is trying to occupy a larger portion of that workflow and make its platform harder to replace.
Strategic AI Positioning
The announcement arrives at a moment when frontier AI and machine learning products are being judged less by model capability alone and more by the quality of their implementation. In enterprise software, the winning products are likely to be those that translate raw AI power into predictable, auditable, and collaborative systems. Figma's latest features appear aimed squarely at that market reality.
The company is not just adding intelligence to design; it is attempting to shape the rules by which intelligence operates inside design. That is a subtle but important distinction. In an environment where agents can draft, modify, and potentially deploy work, the platform that controls permissions and workflow logic may become as important as the model itself.
For customers, the practical appeal is clear: fewer handoffs, faster iteration, and a tighter loop between concept and implementation. For Figma, the commercial logic is equally clear: deeper product stickiness, stronger enterprise relevance, and a more defensible position as AI becomes embedded in creative and technical work alike. The challenge will be execution. Users will expect these tools to be reliable, transparent, and easy to govern, especially as AI moves closer to production environments.
Taken together, the new agent features and the open beta for code layers suggest that Figma is not treating AI as an add-on. It is treating AI as a structural layer in how modern product teams design, collaborate, and build.