- Is it really worth integrating AI into Jira and Confluence for your business?
- What did Atlassian and OpenAI announce, and what does it mean in practice?
- What changes in practice for people who manage projects in a company?
- Three real-world scenarios: how much can an Italian company gain?
- How much does it cost, and what's already included in the plan you pay for?
- What risks and rules do you need to manage before activating AI on your projects?
- Why is everyone talking about it now, and is the Italian market ready?
- How do you get started: the steps to activate Rovo AI in your company
- What are the most common mistakes when activating AI on projects?
- How can you tell if AI in Jira and Confluence is really working?
- So, is it worth it for your company?
- How do we see it at SHM?
- Frequently Asked Questions
- Atlassian and OpenAI have expanded their partnership to connect OpenAI's advanced AI models with Jira, Confluence and Jira Service Management, automating planning and business knowledge management.
- Rovo, Atlassian's AI engine, has been included at no additional cost in all paid Cloud plans since April 2025 and already has over 3 million monthly active users.
- For Italian companies already using Atlassian, the starting point is to activate Rovo and the MCP connector for ChatGPT: no extra investment, just a setup to complete.
- 62% of large Italian companies already use at least one process automation technology, but only 8% have a strategic vision: the risk is adopting the tool without a plan.
Agentic automations on Rovo, Atlassian's AI engine built into Jira and Confluence, have grown 7x in six months. If your company already uses these tools, AI is already inside: the question is whether you're really using it.
Is it really worth integrating AI into Jira and Confluence for your business?
If your company already uses Jira or Confluence on a paid Cloud plan, AI is already included in the price: the question isn't whether to spend, but whether it's worth setting up. If you don't use these tools yet, the evaluation is more nuanced.
| Your situation | Is it worth it? | why |
|---|---|---|
| You already use Jira/Confluence on a paid Cloud plan | Yes | Rovo AI is included at no additional cost: just activate and configure it |
| You use Jira/Confluence on-premise or the Server version | No, for now | Rovo and the OpenAI integration are only available on Cloud |
| You don't use Atlassian but you're looking for a PM tool with AI | Currency | Compare alternatives: the Atlassian ecosystem is mature, but it's not the only option |
| You have distributed teams with a lot of documentation in Confluence | Yes | Rovo Search and AI agents cut the time wasted looking for information |
| You have fewer than 10 people and simple processes | Currency | The gain is real but needs setup time: start with a small use case |
What did Atlassian and OpenAI announce, and what does it mean in practice?
Atlassian and OpenAI have expanded their partnership to connect OpenAI's AI models, the ones behind ChatGPT, with the company knowledge held in Jira and Confluence, so teams can plan, build and deliver work faster.
At the heart of the deal is the Atlassian Rovo MCP connector (Model Context Protocol, a protocol that lets an AI model read and act on external systems). Available since December 2025 and stable since February 2026, it brings the context of Jira, Confluence and Compass directly into ChatGPT. You can create a Jira task, start a workflow or search for a document without leaving the chat.
Put simply: instead of jumping from one tool to another, the AI reads your project, understands the context and takes action. It's not a generic chatbot: it knows your tickets, your deadlines, your documentation.
| Feature | What it does | Who benefits most |
|---|---|---|
| Rovo Search | Searches Jira, Confluence and connected apps using natural language | Teams with a lot of scattered documentation |
| Rovo Agents | AI agents that carry out repetitive tasks on their own (e.g., updating statuses, assigning tickets) | Project managers and development teams |
| MCP connector for ChatGPT | Creates issues and starts Jira workflows directly from ChatGPT | People who already use ChatGPT as a daily work tool |
| AI in Jira Service Management | Suggests replies, classifies requests, reduces ticket handling time | IT departments, internal customer service |
What changes in practice for people who manage projects in a company?
The most tangible change is less time spent on coordination tasks: updating statuses, searching for documents, routing requests. These tasks can be delegated to AI agents, freeing the team for work that requires judgment.
- 1 · Project managerAssisted planningAsks the AI to create a sprint plan from a natural-language description. The AI reads the open tickets and suggests priorities.
- 2 · Development teamAutomatic status updatesRovo agents update ticket status when they detect that a task is complete, so developers don't have to do it manually.
- 3 · IT department / Service deskTicket classification and repliesThe AI classifies incoming requests, suggests replies based on Confluence documentation and cuts handling times.
- 4 · Knowledge managerDocumentation searchRovo Search answers natural-language questions by searching all connected Confluence documents, even ones from years ago.
- 5 · Manager / LeadershipAutomatic reports and summariesThe AI produces project status summaries and highlights delays and blockers, so nobody has to write reports by hand.
Three real-world scenarios: how much can an Italian company gain?
The gain depends on how many people use Jira and Confluence, and how much time they currently spend on repetitive coordination tasks. Here are three illustrative scenarios.
Illustrative examples: numbers are meant to show the reasoning, they are not real customer data.
| Company type | Current situation | What changes with Rovo AI | Estimated weekly time savings |
|---|---|---|---|
| E-commerce (20 people, product + IT team) | The PM manually updates statuses and writes a report every Friday: 3 hours/week | Rovo generates the report and updates statuses automatically | ~2.5 hours/week for the PM |
| Professional firm (10 people, Confluence as a wiki) | Team members search documents for procedures: 30 min/person/day | Rovo Search answers in natural language, finding the right document in seconds | ~1 hour/person/day |
| Manufacturing company (50 people, Jira Service Management for internal IT) | IT manually sorts 40 tickets/week and answers questions that are already documented | AI classifies and answers standard tickets, while IT handles only the complex cases | ~5 hours/week for the IT team |
How much does it cost, and what's already included in the plan you pay for?
Rovo is included at no extra cost in all paid Atlassian Cloud plans (Standard, Premium and Enterprise) for Jira, Confluence and Jira Service Management, starting April 2025.
Here's what many companies don't realize: if you already pay for an Atlassian Cloud plan, you're already paying for AI. You don't need to approve a new budget or run a tender. You just need to turn on Rovo and set it up.
According to Atlassian, after Rovo was included in Cloud plans, AI usage across its customer base has more than doubled. The signal is clear: companies that discover it use it.
| Atlassian Cloud plan | Rovo included? | MCP connector for ChatGPT | Notes |
|---|---|---|---|
| Free | No, for now | No, for now | Basic features only, no AI |
| Standard | Yes | Yes | Rovo Search and basic agents included |
| Premium | Yes | Yes | Advanced automation features |
| Enterprise | Yes | Yes | Additional governance and security controls |
What risks and rules do you need to manage before turning on AI for your projects?
There are three main risks: the quality of the data the AI works with, access governance, and GDPR compliance for employee and customer data.
AI is only as good as the data it finds. If Confluence is full of outdated documents or Jira has poorly structured tickets, Rovo will give inaccurate answers. Before turning on any agent, it's worth spending an hour cleaning up your most-used documentation.
On privacy, data moving between Atlassian and OpenAI is subject to the policies of both companies. If you handle sensitive data (healthcare, legal, finance), you need to check where the data is processed and whether your existing contracts cover this flow. Atlassian offers data residency options in Europe for Enterprise plans.
To learn more about governing AI agents in your company, the article on autonomous AI agents and team structure offers a useful look at oversight risks.
| Risk | Probability | How to mitigate |
|---|---|---|
| Outdated documentation that leads to wrong answers | High | Run a quick Confluence audit before activation: archive documents that aren't up to date |
| Uncontrolled access to AI agents | Medium | Set permissions by role: not everyone should be able to launch automations |
| Sensitive data exposed to external models | Medium | Check Atlassian's data residency policies and OpenAI's terms for your plan |
| Over-reliance on AI for critical decisions | Low | Keep human review for decisions that affect customers or budget |
Why is everyone talking about it now, and is the Italian market ready?
AI adoption among Italian businesses has grown fast, but it's still below the European average: there's a window of advantage for those who move today.
Istat data published in December 2025 show a sharp acceleration: adoption has tripled in two years. But the figure that matters most for decision-makers comes from the Politecnico di Milano: 62% of large companies already use automation, yet only 8% have a strategic vision. Most switched the tool on without a plan. That's the real risk.
The picture is similar across Europe: according to Eurostat, 20% of EU businesses used AI technologies in 2025, with large companies at 55% and small ones at 17%. The gap between those with a strategy and those without is set to widen.
How do you get started: the steps to activate Rovo AI in your company
The most effective path starts with a small, measurable use case, not a company-wide rollout. Here are the steps, in order.
- 1 · IT lead or PMCheck your active Cloud planCheck whether your Atlassian plan is Standard, Premium or Enterprise on Cloud. If so, Rovo is already included: go to the admin settings and turn on Rovo.
- 2 · IT leadRun a quick Confluence auditIdentify the 10-20 most-viewed pages. Update them or archive the outdated ones: Rovo Search is only as good as the documents it finds.
- 3 · PM or team leadChoose a pilot use caseStart with a repetitive, measurable process: IT ticket classification, sprint status updates, or documentation search. Measure the time before and after.
- 4 · IT leadSet up the MCP connector for ChatGPTIf your team already uses ChatGPT, turn on the Atlassian Rovo MCP connector by following the instructions in the <a href=’https://help.openai.com/en/articles/11391654-chatgpt-business-release-notes’ target=’_blank’ rel=’noopener noreferrer’>ChatGPT Business release notes</a>. It lets you create Jira issues straight from the chat.
- 5 · ManagementDefine your governance rulesDecide who can turn on automations, what data the AI can process, and how agent output is reviewed. Put the rules in writing before you scale.
What are the most common mistakes when turning on AI for projects?
The most frequent mistakes aren't technical: they're about method. Companies turn on the tool without defining what they want to measure and without preparing the data the AI has to work with.
- Turning everything on at once. Switching on all Rovo agents without a controlled pilot creates confusion and pushback in the team. Start with a single process.
- Not cleaning up Confluence first. If your documentation is a mess, Rovo Search returns inaccurate answers. AI amplifies the quality of the data it finds, good or bad.
- Not measuring the “before.” If you don't know how much time a process costs today, you can't prove the value of AI afterward. Measure before you activate.
- Ignoring team training. AI agents change the way people work. Without basic training, the team keeps doing things the old way and the tool goes unused. The topic of managing AI tools in enterprise workspaces also offers useful guidance for the internal adoption phase.
- Confusing the tool with the strategy. Rovo is a means, not a plan. As the analysis of enterprise AI agents for customer service points out, value comes when AI is built into a redesigned process, not layered on top of an old one.
How do you know if AI in Jira and Confluence is really working?
Three indicators are enough to tell whether the time you invested in setup is producing concrete results.
| Indicator | How to measure it | Positive signal |
|---|---|---|
| Average response time to IT tickets | Compare the average time before and after you turn on Rovo in Jira Service Management | A reduction of 20% or more in the first 4 weeks |
| Rovo Search usage | Check Rovo analytics: how many searches per day, how many active users | Steady growth over the first 8 weeks: your team is using it |
| Time spent on manual updates | Ask your PM to track the time spent updating statuses and writing reports for 2 weeks before and 2 weeks after | A reduction of at least 1 hour/week per person |
| Confluence documentation quality | Measure the number of pages updated in the last 90 days before and after activation | An increase: AI encourages the team to keep documents up to date |
| MCP connector adoption | How many users have connected ChatGPT to Jira and use it at least 3 times a week | More than 50% of the tech team within 60 days |
So, is it worth it for your company?
Worth it if you already use Jira or Confluence on a paid Cloud plan: Rovo is already included, the marginal cost is zero, and the potential savings on repetitive tasks are real. Start with a 2-week pilot.
Somewhere in between if you're evaluating Atlassian as a new project management tool: treat the AI ecosystem as a plus, but don't choose the platform for that reason alone. First, assess whether your team will actually adopt Jira and Confluence in their daily workflow.
Not yet if you have small teams with simple processes and no structured documentation: in the short term, setup time outweighs the benefit. Wait until you have at least 10 people and a Confluence with real content.
How do we see it at SHM?
For us, the Atlassian-OpenAI integration is an example of how AI should enter a company: not as a new tool to buy, but as a layer of intelligence added to the tools you already use. We believe AI should be integrated to free people from routine work — updating statuses, searching for documents, sorting tickets — and give back time for work that calls for human judgment.
We've worked with companies whose Confluence was full of valuable documentation that nobody could access: no one could find it, no one updated it. When you structure your information architecture before turning on any AI, results are far better. The tool amplifies what's already there: if the foundation is solid, it works. If it's chaotic, it creates chaos faster.
If you're evaluating how to bring AI into your business processes, the starting point is understanding which processes really cost you time today. From there, you build a plan. You can explore our approach to strategic AI integration, process automation training or data processing and business intelligence. If you'd like to talk it through directly, you can also write to us.
For a broader picture of how AI agents are changing team structures, read our analysis of autonomous AI agents and organizational models and follow the AI and process automation cluster for updates.
Frequently Asked Questions
Rovo is included at no additional cost in all paid Atlassian Cloud plans (Standard, Premium and Enterprise) for Jira, Confluence and Jira Service Management, starting April 2025. If you're on the Free plan or a Server/Data Center version, Rovo is not available.
The MCP (Model Context Protocol) connector lets ChatGPT read the context of Jira and Confluence and act on it: you can create a Jira task, start a workflow or find a Confluence document right from the ChatGPT chat, without switching screens. It has been available since December 2025 and became stable in February 2026.
Data moving between Atlassian and OpenAI is subject to the policies of both companies. If your company handles sensitive data, you need to check where the data is processed and whether your contracts cover this flow. Atlassian offers data residency options in Europe for Enterprise plans.
The most effective starting point is a small, measurable use case: automatic classification of IT tickets, searching Confluence documentation, or updating sprint statuses. Before you activate any agent, run a quick audit of your most-used Confluence documents and update or archive the outdated ones.
With a well-defined pilot, the first signs show up in 2-4 weeks: faster ticket response times, more searches on Rovo Search, fewer manual updates for the PM. The full value shows after 60-90 days, once the team has adopted the new workflow.
In March 2026, Atlassian announced layoffs affecting about 10% of its workforce to redirect investment toward AI and enterprise sales. It's a sign of strategic restructuring, not crisis: the company is concentrating resources on AI. Rovo already has over 3 million monthly active users, a figure that points to real adoption.
The tool is already in the plan you pay for: the real work is deciding which process you want to stop doing by hand.
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.
