Team 26 Europe--Atlassian Corporation (NASDAQ: TEAM), a leading provider of AI collaboration and productivity software for teams, announced AMP: The Agentic Multiplayer Protocol at its Team '26 Europe conference. Other announcements include a new Atlassian MCP Server, enhancements to the Teamwork Graph, and capabilities that allow customers to collaborate with agents in richer, more visible ways.

The most significant shift in human-AI collaboration is now underway: a move from isolated AI interactions toward integrated, multiplayer teamwork. Today, work with agents is mostly happening one-on-one and locked away in disconnected silos, like browser tabs or terminal sessions. As AI shifts to multiplayer mode, agents operate in the flow of work, collaborating alongside teammates in the open. This keeps an agent’s work visible to the whole team, preserves context, and turns AI into a true participant in how work gets done.

AMP establishes the foundational rules of engagement for this shift, ensuring that agents have the necessary context, identity, and governance to work safely and effectively as part of a team.

• In-the-Flow Collaboration: AMP lets agents participate wherever work happens, both on an off the Atlassian platform: through @mentions in Confluence, Jira comment threads, or Loom video briefs, while the Atlassian MCP Server pulls in context from third-party tools like Figma or IDEs.
• Identity and Shared Presence: Every agent under AMP has a clear owner and distinct profile. Agents now appear in real-time presence bars and cursors alongside human teammates on canvases and whiteboards. Local agent sessions on developer laptops are bridged directly into Jira boards for full team visibility.
• Context and Governance: Agents are grounded in the Teamwork Graph, drawing on the context relevant to your work. Across Atlassian, the graph now connects more than 250 billion objects and relationships. AMP enforces strict permissions, allowing agents to "Run as User" or use dedicated service accounts, with a full audit trail of every action taken.

"The best work has always been done in teams, and that’s true of AI too," said Mike Cannon-Brookes, CEO and Co-founder, Atlassian. "The companies that pull ahead in the AI era will be the ones that get their people and agents working together in the flow, out in the open, as one team. AMP is how Atlassian is building that future for teams.”

From the keynote stage of Team 26 Europe

Atlassian has shipped over 120 enterprise capabilities during the last year, with even more launched at Team 26 Europe. All innovations build on Atlassian's platform for human + AI collaboration, where people and agents now work together more than 10 million times a month. Highlights include:

A new Atlassian MCP (Model Context Protocol) Server that connects external AI tools and coding agents to enterprise context across Jira, Confluence, Loom, and Bitbucket. Atlassian MCP now has nearly 2 million monthly active users and handles over 15 million daily tool calls — a 15x jump in six months. The new version covers a broader surface while running leaner, using up to 25% fewer tokens for the same Jira and Confluence work (in internal benchmarking on Claude models). Read the blog for more details.

The Teamwork Graph is getting even richer with new context types flowing into it. The Teamwork Graph now reads source code right down to the function, symbol, and class level, with the new Code Search app. Structured data is a native context type now, so users can query data lakes and warehouses from anywhere with Atlassian Insights. Its library of 80+ Connectors continues to grow with ones added from Zoom, Gong, Microsoft Entra ID, and Google Identity. And, Artifacts gives agent outputs a permanent, governed home that indexes straight back into the Teamwork Graph. Read this blog for everything new in Rovo and the Atlassian platform.

New ways to see what agents are doing. As AI adoption grows, work increasingly happens where the rest of the team can't see it, such as in terminals, local sessions, and third-party bots. Atlassian makes that work visible, traceable, and governable. For example, Agent Sessions surfaces cloud and local agent work in Jira and the Teamwork Graph, so context carries forward instead of getting lost in terminals, while Non-Human Identities gives every agent - Rovo or third-party - its own identity, so you always know which agents are active, what they can access, and who owns them.

New ways to collaborate with agents, shifting from one-directional prompting to working alongside them in the open. With Record for Agent, a quick screen-and-voice recording becomes an actionable brief that an agent can turn into Jira work items, a working prototype, or a detailed Confluence spec; and with Interactive PR Reviews, agents record a Loom walkthrough of their own code changes, explaining what they did and the key tradeoffs, so human reviewers can judge the work fast right in Bitbucket.