Atlassian Stock Is Down 84% From Its Peak — And the CEO Just Laid Off 1,600 People to Bet on AI

A certain type of corporate announcement makes a concerted effort to project confidence while subtly acknowledging that things aren’t going as planned. Mike Cannon-Brookes released one of those in March 2026. The co-founder and lone CEO of one of Australia’s most renowned tech companies stated in a blog post that Atlassian would lay off about 10% of its workforce, or 1,600 jobs, in order to “self-fund further investment in AI and enterprise sales.” The wording is deliberate. Even thoughtful. However, the pressure beneath it is not completely hidden by it.

This year, Atlassian’s stock ticker, TEAM, has taken on a somewhat unsettling symbolic meaning. The company, which produces the Confluence wiki tool used by hundreds of thousands of businesses worldwide and Jira project-tracking software, has lost more than half of its value in 2026 alone. It is currently about 84% below the record high it reached during the pandemic-era software boom of 2021. Atlassian felt like one of the most obvious beneficiaries of a world reorganizing around remote work back then, when office workers were confined to their homes and demand for cloud collaboration tools was at an all-time high. Investors priced the Covid tailwind appropriately because it was real. Then the wind changed.

What followed wasn’t exclusive to Atlassian; the emergence of generative AI has forced the entire enterprise software industry to undergo a challenging reevaluation. The worry that tools like Jira and Confluence occupy a layer of the software stack that AI agents may soon render redundant, or at least drastically cheaper to replicate, is expressed in analyst reports, earnings calls, and pretty much everywhere else. It’s still unclear if that worry is entirely warranted. However, markets don’t wait for clarification. Shares began to decline and continued to do so.

It’s difficult to ignore the fact that Cannon-Brookes is, in a way, attempting to outrun a narrative that the market has already written about his business. The layoffs, which are the second in three years after a 500-person reduction in 2023, are presented as a shift toward artificial intelligence rather than a retreat. As of February, Atlassian boasted five million monthly users of Rovo features. The company has been developing its Rovo AI platform and has launched an open beta for AI agents integrated into Jira and Confluence. For three consecutive quarters, the company’s year-over-year revenue growth has actually accelerated. This is the kind of information that can be obscured by all the noise surrounding the stock price. Growth in revenue is one thing, but investor confidence in the long-term business plan is quite another.

Fundamentally, the Atlassian story is a truly improbable one. In 2002, Cannon-Brookes and Scott Farquhar launched the business in Sydney using $10,000 in credit card debt after Cannon-Brookes asked his university classmates via a mass email if anyone was interested in starting a startup. Only Farquhar responded. The Bronze Statue at Rockefeller Center in New York is said to have inspired their choice of name, which is derived from the Greek mythological figure Atlas. As pure an origin story as the tech industry has ever produced, they created Jira in part because they were dissatisfied with the bug-tracking software they were already using. They became Australia’s first tech startup billionaires when they went public in December 2015, despite The New York Times, with its trademark dry wit, referring to Atlassian as a “very boring software company.”

For a long time, that dullness was more of a feature than a flaw. By definition, enterprise software is unglamorous. Just as an office building depends on its plumbing, tens of thousands of engineering teams rely on Jira on a daily basis without giving it much thought. Deep operational dependency of that kind is meant to be a moat. Because they have trained entire teams on the product, their workflows are centered around it, and their institutional knowledge resides within it, customers are reluctant to switch. With a cloud-first migration strategy, Atlassian capitalized on this stickiness by encouraging users to switch from legacy “Server” editions, for which it stopped providing support in early 2024, to Cloud and Data Center subscriptions. The story that investors bought was simple: steady recurring income, growing profit margins, and a strong competitive position.

The sheer speed at which software teams are evolving now obscures that story. A real wager on the direction of the enterprise market is Atlassian’s Rovo and MCP ecosystem, a collection of AI agents made to function with Jira, Confluence, and third-party tools. Whether Atlassian can meaningfully monetize those capabilities or if it is creating features that consumers will expect for free while larger platforms absorb the underlying demand is the central question in the current investment debate. By 2028, analysts predict the company will generate $8.7 billion in revenue, which would require an annual growth rate of about 18.7%. That is doable. Additionally, there is no guarantee.

Farquhar, who resigned as co-CEO in August 2024 but is still on the board, left at an intriguing time—not when the business was in crisis, but rather when things were obviously changing. Now in charge alone, Cannon-Brookes is managing a $225 to $236 million restructuring charge while maintaining that AI is changing Atlassian’s skill mix rather than just replacing employees. “It would be disingenuous to pretend AI doesn’t change the mix of skills we need,” he stated. In contrast to some of the more euphemistic AI announcements that have recently swept Silicon Valley, there is something almost refreshingly honest in that statement.

Observing Atlassian’s trajectory gives the impression that the company’s struggles are more about the awkward renegotiation that every established software company must undergo with itself than they are about failure, at least not yet. Atlassian’s indispensable tools were designed for a world in which every workflow, bug, and decision was managed by humans. While AI doesn’t completely erase that world, it does begin to question whether the software that runs on top of it needs to have the same appearance. With a lot of uncertainty and a decline from its peak, Atlassian stock is essentially a bet on the company’s ability to respond to that question. Longer odds have previously faced the founders who began with $10,000 and an unanswered email.