Canva has launched Canva AI 2.0, a major expansion of its AI offering that positions the platform as a more conversational and agentic workspace for design, documents, presentations and broader workplace tasks.
Announced at Canva Create in Los Angeles, the update introduces a new architecture layer and a suite of AI-powered workflows, with the company rolling out the research preview to one million users starting April 16.
What is Canva AI 2.0?
At the centre of the launch is a shift from one-off AI generation to a system designed to accompany users throughout the creative process. Canva said the platform can now move from idea to publish in a single conversation, while keeping outputs editable and collaborative across formats.
In a LinkedIn post following the launch, Cameron Adams, co-founder and chief product officer at Canva, described the release as “Canva reimagined as a conversational, agentic platform”. He said the company had spent more than a decade building the foundations for the product, and more than two years making those foundations accessible to AI.
Adams said the goal was not simply faster generation, but making AI more useful across the entire creative process. “AI is most powerful when it augments human creativity, not replaces it,” he wrote.
New features include connectors, scheduling and web research
Canva AI 2.0 includes conversational design, agentic orchestration, object-based editing and what the company calls “living memory”, which adapts to a user’s style, brand and working preferences over time. It also adds new workflow tools, including connectors for Slack, Notion, Zoom, Gmail, Google Drive and Google Calendar.
Other additions include scheduled background tasks, built-in web research, brand intelligence, updated Canva Code tools with HTML importing, AI-powered spreadsheet generation and an expanded template remix feature. The release is designed to extend Canva’s role beyond visual creation to include planning, research, and workplace collaboration.
Canva highlights in-house AI research
Canva also used the launch to spotlight its internal AI research team. According to the company, more than 100 researchers across Canva Original Research and Exploration are building multimodal foundation models designed specifically for design workflows.
The company said its proprietary models are up to seven times faster and 30 times cheaper than comparable frontier alternatives, with model deployment cycles reduced from more than two years to as little as a month. On page nine of the release, Canva said these efforts now span image, video, style transfer and structured design generation.
Claude and ChatGPT integrations expand Canva’s AI reach
As part of the update, Canva said it has deepened its collaboration with Anthropic, bringing Canva’s Design Engine and Visual Suite into Claude. Users can also import artefacts from AI platforms, including Claude and ChatGPT, into Canva, where they become editable and collaborative assets.
Canva said this reflects a broader strategy to embed its platform across more AI-driven workflows. The company also cited Andreessen Horowitz research claiming that Canva is now the third-most-used AI platform globally and the fastest-growing in customer spend on AI products among leading software companies.
Top image: Cliff Obrecht, Melanie Perkins and Cameron Adams