The rapid rise of generative AI has dramatically increased the volume of visual content being created every day. However, much of that content comes with a frustrating limitation: once generated, the design is often locked inside a static image file.

Even small edits—changing text, repositioning elements, or adjusting layouts—can require starting the creative process over again.

To address this challenge, Canva has introduced Magic Layers, a new AI-powered technology designed to transform flat images into fully editable design files.

The feature, now available in public beta, allows users to convert static images into structured layers that can be edited directly within the Canva editor.

The result is a workflow where AI-generated visuals serve as a starting point for design rather than a finished, unchangeable output.

Unlocking Static Images

Traditional image formats such as PNG or JPG store designs as flattened visuals. Once exported, the original structure of the design disappears.

Text becomes pixels, shapes merge together, and the relationships between design elements are lost.

That means editing an image often requires rebuilding it from scratch.

Magic Layers aims to reverse that process.

By analyzing the structure of a flat image, the technology identifies separate components within the design and reconstructs them as editable elements.

These elements are then placed into layers within Canva’s editor, allowing users to move, modify, or replace them as if they were working with the original design file.

From Prompt to Editable Design

Magic Layers also works with AI-generated designs created inside Canva.

Instead of producing a static image, the system generates designs that remain fully editable from the beginning.

Users can create visual content from a prompt and immediately refine it by:

• Moving design elements

• Changing fonts or text content

• Replacing backgrounds

• Adjusting layout positioning

• Customizing colors and styles

This approach eliminates the need to repeatedly generate new images when a design requires small adjustments.

Instead, creators can iterate directly within the design environment.

How Magic Layers Works

The feature begins by analyzing the structure of an image.

When a user uploads a flat image, the system performs several actions automatically:

• Separates visual elements into individual movable objects

• Restores text as editable text boxes

• Preserves layout relationships between design components

• Maintains the original visual structure

The result is a design that closely matches the original image but behaves like a fully layered file.

For designers and marketers, this means existing images can be transformed into scalable assets without recreating them manually.

Beyond Traditional Vector Tracing

Tools capable of converting raster images into vector shapes have existed for years, but they come with important limitations.

Traditional vector tracing tools focus on identifying shapes and converting pixel regions into outlines. While this can reproduce visual forms, it does not capture the meaning or relationships between elements.

For example, a tracing tool cannot determine whether a shape represents:

• A background object

• A text block

• A design element grouped with others

Magic Layers approaches the problem differently.

Instead of simply tracing shapes, it analyzes the entire design structure to interpret how elements relate to each other.

This allows the system to restore editable text, maintain alignment relationships, and preserve the overall layout of the design.

The result is not just a traced image, but a reconstructed design file.

Built on Canva’s AI Design Model

Magic Layers is powered by the Canva Design Model, the company’s proprietary AI foundation model designed specifically for visual communication.

Since its introduction in 2024, the model has generated hundreds of millions of editable assets across formats such as:

• Presentations

• Documents

• Social media posts

The model also powers Canva’s integrations with major AI ecosystems including:

• ChatGPT

• Claude

• Microsoft Copilot

With Magic Layers, Canva extends the model’s capabilities beyond generating designs to reconstructing existing ones.

A New Workflow for Creators

For creative teams, the biggest value of Magic Layers may lie in how it changes the workflow around AI-generated content.

Generative AI has made it easier than ever to produce visual ideas quickly. However, the inability to edit those outputs has often limited their practical usefulness.

Magic Layers turns that process into a more flexible creative loop.

Instead of treating AI-generated images as final products, users can treat them as drafts that can be refined and adapted to different contexts.

For example:

• Marketing teams can modify AI-generated visuals to match brand guidelines

• Small businesses can update messaging without redesigning assets

• Content creators can remix visual concepts into new formats

This shift moves AI-generated content from a “one-shot” process to an iterative design workflow.

Beta Availability and Future Expansion

Magic Layers is currently available in public beta and supports single-page PNG and JPG files.

The feature is rolling out initially in:

• The United States

• The United Kingdom

• Canada

• Australia

Canva plans to expand support for additional file types and design capabilities in future updates.

The Bigger Picture

The launch of Magic Layers reflects a broader trend in AI-powered creativity.

While generative AI tools have made it easier to produce visual content, creators still need control over the final design.

Editable AI outputs represent a significant step toward combining automation with traditional design workflows.

By enabling users to modify AI-generated images without starting over, Canva aims to bridge the gap between generative AI and real-world creative production.

For designers, marketers, and everyday creators, the message is clear: AI may generate the first version of a design, but the creative process doesn’t end there.

Get in touch with our MarTech Experts.