June 8, 2026
There’s a problem that most organisations using AI tools aren’t talking about openly: individual productivity has jumped, but company-wide output hasn’t kept pace. People are getting faster in their own lanes, but the work still fragments when it comes together.
Miro has made this the central problem its latest platform updates are designed to solve. At Canvas 26, its annual customer event held in San Francisco on 19 May 2026, the collaboration platform announced a series of updates to its AI workspace that aim to bring human-to-human collaboration, human-to-agent workflows and agent-to-agent processes onto a single shared surface. The argument is that operating in silos turns AI into a misalignment amplifier rather than a collaborative solution.
Andrey Khusid, CEO and Founder at Miro, stated the problem plainly: “AI leverage is locked inside private chat windows – accelerating individuals, but never reaching the organisation. When every collaboration mode converges on one surface, individual speed becomes company speed, and individual clarity becomes shared clarity.” Matt Cloke, CTO at Endava, added the practitioner’s version: “Accelerating work with AI in a silo creates speed without direction – and that’s a problem.”
The headline update is that Miro’s canvas can now be read and written by third-party agents.
This means external AI agents can contribute directly to shared boards rather than working around them. Miro has expanded its MCP support to cover tool and board creation, frames, comments, shapes and code blocks and added agent-friendly formats including Mermaid diagrams, Markdown and HTML widgets. New Connectors link Miro’s own tools to Slack, Atlassian, Granola, GitHub and others, while also embedding it natively within ChatGPT, Claude and Microsoft Copilot.
Sidekicks, Miro’s agentic AI assistant, has been upgraded to move beyond reactive prompt-response behaviour. It can now understand intent, break ambiguous problems into steps, ask clarifying questions and generate full board content from a single prompt – documents, diagrams, Kanbans, sticky notes and frames. The addition of persistent memory means Sidekicks builds context over time, picking up where it left off. Voice interaction has also been added, allowing two-way conversation rather than typed prompts.
Flows, Miro’s automated workflow tool, now extends beyond the canvas through the new Connectors, pulling in meeting transcripts, creating tasks in project trackers and surfacing Kanban views across connected systems. The update includes human-in-the-loop approval steps, positioning Flows as shared infrastructure for repeatable team work rather than individual automation.
Miro Prototypes received updates focused on reducing the distance between idea and team alignment. It can now pull context directly from Claude Code, import screenshots or Figma files to generate editable multi-screen flows, apply brand styling automatically from a URL, generate multiple variants simultaneously and export directly to coding agents or Figma for handoff.
Taken together, the updates reflect a consistent concept: that the bottleneck in enterprise AI adoption isn’t individual capability, it’s shared context. Whether that argument resonates will depend on whether teams find the canvas a natural place to bring AI output together – or whether the silos simply move to a different tool.
June 30, 2025
April 10, 2026
April 15, 2026
[email protected]
TechRound
83 Charlotte Street
London W1T 4PR
Download our media pack
We are delighted to announce that award submissions for 2022 are open. We are running the Techround 100, BAME 50 and the 29 under 29 competitions.
Enter now and get your business in front of an influential audience of over 300,000 monthly visitors.
No thanks, I’m not interested!
Leave a Reply