Canva has launched AI 2.0, a system that creates editable, multi-layered designs from conversational prompts using a foundation model built specifically for design tasks. The move positions the Australian design platform to compete in the generative AI era.
Canva's new AI 2.0 tool represents a shift in how the platform approaches design automation. Rather than generating static images, the system produces designs with separate, editable layers—allowing users to modify individual elements without regenerating the entire composition.
The foundation model underlying AI 2.0 was built specifically for design work, distinguishing it from general-purpose generative AI tools. This specialization aims to deliver outputs that better understand design principles, layout, and the relationships between visual elements.
Users interact with the tool through conversational prompts, reducing the friction between concept and execution. A designer can describe what they want in natural language, and the system translates that into organized, workable designs.
The layered output proves crucial for practical design work. Traditional AI image generators produce finished outputs that resist modification. Canva's approach maintains the flexibility designers need—if a generated element doesn't fit, users can edit it directly without starting over.
Canva, valued at $26 billion, has built its business on democratizing design tools. The company claims 200 million monthly active users. AI 2.0 extends this mission by automating the initial design phase while preserving user control.
The unveiling reflects competitive pressures in the market. Design platforms face pressure from general-purpose AI tools that can generate images from text. Canva's response prioritizes the professional design workflow over raw generation capability.
The foundation model approach also signals a broader trend: as generative AI matures, specialized models tailored to specific domains are gaining prominence. Rather than relying on OpenAI or other third-party APIs, companies are building proprietary models optimized for their use cases.
Canva's move carries execution risk. Delivering AI that truly understands design workflow and produces immediately usable assets remains technically challenging. The test lies not in novelty, but in whether designers actually adopt AI 2.0 for their daily work.
Z.ai released GLM-5.3's weights on Hugging Face under a new license that requires large companies to undergo security review before hosting the model. The change marks a departure from the standard MIT license.
Anthropic has introduced the Model Hardware Standard (MHS), a unified interface enabling AI agents to operate robotic arms, lab instruments, and other physical devices. Early testing shows integration time has dropped from weeks to hours.
Open-weight AI companies—those releasing freely available models—are attracting major acquisition interest from tech giants. The trend reflects growing capital investment in the business model of distributing AI models at no cost.
Google Deepmind has upgraded its Co-Scientist AI system to autonomously plan experiments, operate lab equipment, and publish scientific papers. The Gemini-based multi-agent platform demonstrated experimentally validated results across materials science, chemistry, and medical AI development.