Atlassians cloud rebound|A cycle or a reset?

A stronger cloud transition

Atlassian has produced the kind of quarter that changes the AI question, but does not settle it. Revenue rose 28% to US$1.766 billion, cloud revenue accelerated 31%, and the company delivered its first quarterly operating profit in more than two years.

The provisional answer is that fears of AI making Atlassian obsolete have been replaced by evidence of a stronger cloud transition. What remains unresolved is whether this is a durable change in customer economics or a sharp recovery from a low expectation base.

Why the AI concern mattered

Earlier in 2026, Atlassian shares had fallen by more than half amid job cuts and concerns that AI tools could make Jira, Confluence and Trello less essential.

The concern was not irrational. Atlassian’s core products depend on teams entering tasks, updates and documentation, while newer AI tools can automate parts of that work.

Growth with operating leverage

The latest results change the timing of that argument. Cloud revenue reached US$1.213 billion, up 31 per cent, after the company had been guiding to roughly 23 per cent cloud growth as recently as February.

Operating income swung from a US$28 million loss to US$211 million profit, while free cash flow reached US$475 million. That combination matters because it shows growth returning alongside operating leverage, rather than growth being bought entirely through higher spending.

Deeper customer commitments

There is also evidence that customers are making deeper commitments. Subscription annual recurring revenue reached US$6.606 billion, up 23 per cent, while remaining performance obligations rose 44 per cent to US$4.817 billion.

Atlassian also reported a record quarter for contracts worth at least US$1 million a year. Those figures suggest the platform is still becoming embedded in large organisations, giving the company more time and data with which to make its AI products useful.

The AI-enabled platform case

Atlassian says that usage of its Model Context Protocol server and Teamwork Graph command-line tool more than doubled to over one million monthly active users.

Rovo, its AI assistant, is now used by more than 80 per cent of Fortune 500 organisations, according to the company. The proposed mechanism is that 25 years of project and collaboration data gives Atlassian context that general-purpose AI tools do not automatically possess. If that context helps agents coordinate real workplace processes, AI could make the platform more valuable rather than replace it.

The relief-rally reading

But there is a less dramatic reading. The shares rose more than 30 per cent because the results beat a market that had already been heavily disappointed, and co-founder Mike Cannon-Brookes plans to buy up to US$250 million of stock.

That buyback may support sentiment, but it does not prove that AI features will become a durable source of revenue.

A narrower growth profile

The company’s own forecast also limits the celebration. Atlassian expects total revenue growth of about 13 per cent in fiscal 2027, cloud growth of 25.5 per cent and subscription ARR growth of 18 per cent.

Data Centre revenue is expected to fall about 17 per cent. So the business is not returning to its old growth profile across every product. It is moving customers from one model to another, and asking the cloud business to carry more of the company.

The reset still has to be earned

For a holder, the question has shifted from whether Atlassian is immediately exposed to AI disruption to whether it can convert AI usage into durable contracts, revenue and cash flow. For a watcher, the important distinction is between a relief rally and an operating reset. The next useful checkpoint is whether future fiscal 2027 results hold close to the cloud and ARR guidance while contracted revenue continues to grow.

The strongest current judgement is that Atlassian has demonstrated resilience and a credible transition into AI-enabled cloud software, not yet a proven structural moat. The available evidence does not establish how much of the reported AI usage is paid, how profitable those features will become, or whether the slower group-wide growth forecast is temporary. That is where the reset still has to be earned.