At today’s Canva Create event, Canva unveiled the research preview of AI 2.0. The upgrade makes the platform a conversational partner that can generate layouts, edit assets and publish finished work from a single prompt. Powered by Proteus, Lucid Origin and I2V models, it runs up to seven times faster and 30× cheaper than comparable tools.

Under the new architecture layer, users describe an idea or brief and the AI assembles a fully editable design with brand‑consistent layout. An orchestration engine selects the right tools, while object‑based intelligence lets creators tweak a single element—image, headline or font—without disturbing the rest. A “living memory” feature absorbs past projects to tailor suggestions over time.

Beyond creation, AI 2.0 adds workflow tools that pull data from Slack, Notion, Zoom, Gmail and Google Drive. It can summarize meeting transcripts, turn emails into customized pitches, or auto‑generate newsletters. Scheduled tasks run offline, producing weekly social batches or briefing docs from calendar events. The revamped Canva Code 2.0 now supports HTML import, turning textual prompts into responsive, interactive experiences.

The preview opens today to the first million users who sign up via Canva’s homepage, with access slated to roll out to broader audiences over the next weeks. By consolidating design, research and automation in one chat‑driven interface, the suite promises to shrink the time teams spend juggling separate apps, positioning Canva as a more complete creative hub.