At Atlassian’s recent Team’26 conference, the company released a broad set of announcements organized around the thesis that organizational context is the primary source of AI differentiation. The announcements clearly show that Atlassian intends to serve as the infrastructure layer that makes that context actionable.

The company opened its Teamwork Graph (its proprietary knowledge network with more than 150 billion connections spanning Jira, Confluence, Loom, and integrated third-party SaaS tools) to external agents and developer tooling via new open interfaces.

It also advanced Rovo from a context-aware AI assistant into an agentic execution platform capable of multi-step, autonomous work across the Atlassian product surface.

The announcements span every major product in the Atlassian portfolio:

• Agents become first-class participants in Jira workflows.
• Confluence gains the ability to host third-party agents and transform structured content into visual formats without leaving the source document.
• Loom adds multimodal agent briefings.
• Rovo Studio, the no-code agent and automation builder, reached general availability for all users.
• Atlassian’s AI-powered browser Dia gained enterprise security certifications and a proactive morning briefing capability.

Underpinning all of it is a deliberate architectural decision to make the Teamwork Graph consumable by any MCP-compatible tool in the enterprise stack.

Atlassian’s platform is no longer primarily a project and knowledge management toolset augmented by AI features. It is instead being aggressively repositioned as an agentic work operating system in which human and AI contributors share a common context layer, participate in the same workflows, and produce auditable, governed outputs.

Announcement Summary

The scope of the announcements is broad, spanning platform infrastructure, product-level capability additions, and developer tooling. The announcements fall into five categories: Teamwork Graph openness, Rovo agentic capabilities, product surface updates across Jira, Confluence, and Loom, new product collections, and enterprise readiness for the Dia browser.

Let’s take a quick look at each.

Teamwork Graph

The Teamwork Graph is Atlassian’s internal knowledge representation layer, mapping people, projects, documents, code repositories, decisions, and relationships across connected tools.

Two new external access mechanisms enter open beta:

• Teamwork Graph CLI: A command-line interface with 300+ commands that allows coding agents (including Anthropic Claude Code and Cursor)  to query and write back to the graph across Atlassian products through a single interface.
• Teamwork Graph tools in Rovo MCP Server: An MCP-standard server that gives any compatible AI client a secure, governed channel to read and act on organizational context. This includes Anthropic’s Claude, Cursor, Google Gemini CLI, Lovable, and WRITER
• Custom Teamwork Graph Connectors via Forge: Now generally available, allowing enterprises and partners to build connectors to proprietary internal systems and legacy platforms. Data ingested through these connectors automatically surfaces in Rovo and Atlassian Analytics, with permissions preserved.

Rovo Agentic Platform

Rovo expands from a retrieval- and chat-oriented AI assistant into a platform capable of autonomously planning and executing multi-step workflows.

Key updates include:

• Max Mode (coming soon, early access): A new reasoning mode in Rovo Chat that accepts complex, open-ended instructions, decomposes them into concrete steps, executes across Jira, Confluence, JSM, and connected tools, and returns structured outputs. It autonomously creates Jira work items, drafts documents, sends notifications, and schedules calendar time.
• Rovo Studio GA: The unified workspace for building agents, automations, and custom apps is now generally available to all users. Studio now accepts a plain-language problem description, recommends whether the right solution is an agent, automation, or Forge app, and then builds it.
• Code Intelligence (early access): Enables engineers and coding agents to ask intent-level questions across multi-repository environments. By combining source graph context with Jira and Confluence data, it answers questions about code ownership, migration status, and architectural patterns rather than conducting simple string searches.
• DX Integration (AI Experience): DX, acquired by Atlassian in 2025, now offers Agent Experience, AI Code Insights, and AI Pulse dashboards that track where AI generates code, how agents perform, and their impact on productivity and reliability.

Jira

Jira receives updates that embed agents directly into project workflows:

• Agents in Jira (GA): AI agents can be assigned work items, @mentioned in comments, and embedded into workflow automations. Actions taken by agents are logged, auditable, and visible alongside human activity. Supported agents include Atlassian’s own Rovo agents and third-party tools, including Amplitude, Canva, Cursor, Figma, Gamma, and GitHub Copilot.
• Create with Rovo in Jira (beta): Converts Confluence documents, meeting summaries, and email threads into structured Jira work items.
• Jira Product Discovery Enterprise (GA): Portfolio-level governance capabilities for product discovery reach general availability.

Confluence and Loom

Confluence and Loom gain capabilities that tighten the connection between content creation and agentic execution:

• Remix with Rovo (beta): Any text-based Confluence content can be transformed into charts, timelines, infographics, geo maps, org charts, quadrants, or flip cards without modifying the underlying source document.
• Third-Party Agents in Confluence (open beta): Users @mention third-party agents from Lovable, Replit, Databricks, and Gamma directly in Confluence pages. Agents read page context and take action across connected tools.
• Agent Briefings in Loom (coming soon): Users record video walkthroughs of requirements, designs, or feedback. Speech, on-screen content, and click context are captured as multimodal input, converted into a structured agent prompt, and can generate Jira work items in a single click.

Product Collection and Dia

Two additional announcements extend Atlassian’s product surface into adjacent categories:

• Product Collection (early access): A new Collection that combines Jira Product Discovery with a new Feedback app, which aggregates signals from support tools, review sites, and demand platforms and connects them to product goals. A Pendo integration provides additional product analytics context.
• Dia Enterprise Readiness: Atlassian’s AI-powered browser now includes layered defenses against prompt injection, SSO, Chromium MDM support, and SOC 2 Type II attestation. A new morning briefing feature surfaces a personalized daily overview, drawn from overnight Slack messages, calendar events, and assigned action items. Dia Reports, a new capability, generates proactive briefings, such as interview preparation documents and decision memos, by combining Teamwork Graph context with everyday browser activity.

Analysis

The most immediate impact of Atlassian’s set of announcements falls on two groups: software development teams and the enterprise administrators who govern Atlassian deployments.

For software teams, the combination of Agents in Jira, Code Intelligence, and Rovo Dev provides a coherent path from issue creation to code execution on the existing tool surface, without requiring developers to adopt new interfaces.

Agent Briefings in Loom take this further, allowing verbal or visual walkthroughs to be structured as work items. This is a workflow shift with material implications for distributed and async teams.

Overall:

• Enterprise administrators gain a more predictable deployment environment through Jira seasonal releases and the governance controls embedded in Rovo Studio GA. All agent actions are logged and scoped to existing permissions and approval flows.
• Product managers who use Jira Product Discovery receive a long-requested capability: the Feedback app provides a direct feedback loop from customer signals to product plans, without manual aggregation.
• Teams using the Atlassian platform in regulated environments benefit from Dia’s SOC 2 Type II certification and MDM support, lowering the barrier to enterprise browser adoption.
• Organizations that rely on proprietary internal systems can now connect that data to the Teamwork Graph via custom Forge connectors, without waiting for Atlassian to build a pre-packaged integration.

Competitive Landscape

Atlassian’s portfolio places it in direct competition with Microsoft Copilot, ServiceNow Now Assist, and Salesforce Agentforce in the enterprise AI platform space. Each competitor has distinct architectural advantages and weaknesses relative to Atlassian’s positioning.

The table below summarizes key comparison dimensions.

Vendor	Org. Context Graph	Agents in Workflow	Developer Experience	No-Code Agent Build	Ecosystem Openness
Atlassian (Rovo)	Teamwork Graph: 150B+ connections across Jira, Confluence, Loom, third-party SaaS	Agents in Jira; MCP open ecosystem	Code Intelligence, Rovo Dev, DX integration	Rovo Studio: agents, automations, apps	MCP Server: CLI open beta, Forge connectors
Microsoft (Copilot)	Microsoft Graph: integrates M365 apps, calendar, email, Teams	Copilot Studio agents; Teams integration	GitHub Copilot (separate product); VS Code integration	Copilot Studio (agent builder, low-code)	Selective; tightly tied to Azure/M365 ecosystem
ServiceNow (Now Assist)	Now Platform data graph: strong in ITSM/workflow context	AI agents in workflows; agentic process automation	Limited native dev tooling	Flow Designer; App Engine Studio	Primarily proprietary; limited external MCP support
Salesforce (Agentforce)	Data Cloud unified customer data; weaker on dev/project context	Agentforce: autonomous sales/service agents	Weak native software dev context	Agent Builder (low-code); Flows	Salesforce ecosystem-centric; MCP support announced

Atlassian’s strongest differentiation relative to these competitors is the depth of its developer and project context. Atlassian’s Teamwork Graph spans the full software development lifecycle and connects to the knowledge layer where product decisions live.

Looking at Atlassian’s most direct competitors:  

• Microsoft Graph is broad but weighted toward communication and calendar data.
• ServiceNow has a strong ITSM context but a limited software development history.
• Salesforce Agentforce is well-positioned for customer-facing workflows but carries minimal organizational knowledge for software delivery teams.
• The MCP openness strategy is a meaningful differentiator relative to Microsoft’s M365-centric posture, and Atlassian’s 93% MCP usage rate among enterprise customers suggests the open ecosystem approach resonates with large customers.
• Microsoft’s breadth of enterprise presence, encompassing identity, productivity, security, and infrastructure, gives Copilot distribution advantages that Atlassian cannot replicate. Where CIOs are consolidating on M365, Copilot’s default availability makes it a low-friction AI entry point regardless of capability depth.
• ServiceNow competes most directly in ITSM and enterprise workflow, where its Now Platform graph is mature and its governance models are well established. Atlassian’s JSM gains from Rovo-assisted incident response, and the Incident Command Center is targeted directly at this overlap.
• Salesforce’s Agentforce announcement earlier in 2026 generated significant market attention, but its weak position in software development context limits its relevance for the engineering and product teams that constitute Atlassian’s core customer base.

The most credible competitive threat to Atlassian’s context graph thesis comes from GitHub Copilot Workspace and the broader GitHub platform, which are controlled by Microsoft.

GitHub holds substantial code context and CI/CD history, and as Microsoft tightens the integration among GitHub, Azure DevOps, and Copilot, it is building an alternative developer context layer that could challenge the Teamwork Graph’s position for engineering teams.

Atlassian’s response to partner with GitHub Copilot as a third-party agent in Jira, rather than compete directly, is pragmatic and acknowledges the overlap.

Final Thoughts

Atlassian is taking its platform in a coherent, well-thought-out direction. The opening of the Teamwork Graph through MCP and CLI interfaces has the broadest strategic reach, converting what was a proprietary context layer into an open infrastructure component that any agent or automation tool can consume.

If this approach gains adoption, the competitive question shifts from which AI assistant is most capable to which organizational context layer is most complete. This is a competition in which Atlassian’s 20 years of enterprise project and knowledge data provide a durable structural advantage.

For enterprise buyers evaluating the Atlassian platform, Team ’26 showed a clear product roadmap and a defensible architectural thesis. Organizations already running significant portions of their software delivery and knowledge workflows on Atlassian tools are well-positioned to extract compounding value as the Teamwork Graph density increases and agentic capabilities mature.

Atlassian’s overall competitive position in the enterprise AI platform market is materially strengthened by these announcements, and the opening of the Teamwork Graph creates a new front in the contest over which vendor’s organizational context layer will become the AI substrate for the modern enterprise.