2 Sources
2 Sources
[1]
Atlassian swears it can deliver AI without blowing out costs
Atlassian has assured investors it can add AI to its services without blowing out its costs or shrinking margins. CEO Mike Cannon-Brookes used the company's Thursday earnings call to reveal Atlassian now has five million users of its Rovo agentic AI offering and suggested that investors in the company might worry that costs would blow out as a result. "We're able to deliver those five million Rovo seats and continue to improve gross margin," he reassured. "That's a huge achievement on behalf of our engineering teams, but it shows that we can manage those AI costs inside for the vast majority of customers." Atlassian runs entirely in public clouds, and Cannon-Brookes' remarks suggest optimizing its operations, rather than high prices, is the reason it's kept margins stable. Which isn't great news for cloud operators, as it suggests at least one SaaS giant has found a way to add AI to its offerings without spending much more. The Australian collaborationware company is, however, sending more business to big clouds by ditching its on-prem datacenter products. CFO Joe Binz said Atlassian "saw very healthy cloud migrations in Q2" and Cannon-Brookes pointed to the company winning $1 billion of revenue from cloud products for the first time in this quarter. The CEO also batted away suggestions that Atlassian, like other SaaS companies, is susceptible to attack by AI. "I'm convinced AI is great for Atlassian. Others think software is dead," he wrote in his shareholder letter [PDF]. " In this environment, it seems that noise swamps signal, nuance gets lost," he added. On the earnings call, he expanded on that point by saying Atlassian customers still want its wares. "When I talk to customers, they believe that we're helping them through a lot of that noise," he said. "Enterprise customers want a platform they can trust. They need it to be compliant and secure and all the things they've always needed." And they're spending more with Atlassian to get those things: Cannon-Brookes said the company continues to see customers adopt more of its stack and commit to using it for longer. He also shrugged off suggestions that Anthropic, which recently previewed a product called "Claude Cowork" that competes with Jira, is a threat because the new AI product will need to access information stored in Atlassian's software. It's not all rosy for Atlassian because Cannon-Brookes declared himself "frustrated" by Atlassian's share price, which has fallen 70 percent in the last twelve months, and fell two more points in after-hours trading. The company is therefore accelerating the pace of stock buybacks. "We believe our shares are significantly undervalued," Cannon-Brookes said. Q2 revenue landed at $1.6 billion, up 23 percent year over year. Operating losses narrowed from Q2 2025's $57.5 million to $47.7 million. ® When The Register started work on this story, we accessed the company's shareholder letter and were a little startled that its opening sentence read "We had a fantastic Q2. We're building a f**king great business." We closed that tab and later loaded the letter again to find it had a new filename and the text had changed to "We're building a bloody great business." Outlets who covered the company before we completed this story record the executive F-bomb.
[2]
Atlassian CEO Says AI Positive for Company as Tech Shares Fall
Cannon-Brookes said the company is continuing to spend significant money on research and development in artificial intelligence, with R&D spending for the quarter at $826 million. AI has been one of the best things to happen to Atlassian Corp., Chief Executive Officer Mike Cannon-Brookes said, as shares fell in the US in late Thursday trading following the company's second-quarter results, amid a broader AI-driven tech sector slump. "There's a lot of noise in the market, there's no doubt about that," he told Bloomberg Television in an interview. "I've been very public, I think that AI is one of the best things that's ever happened to Atlassian, that ever happened to my part of the software industry." Atlassian shares fell more than 8% in late trading in New York and have fallen more than 62% in the year to date. Cannon-Brookes said the company is continuing to spend significant money on research and development in artificial intelligence. R&D spending for the quarter was about 22.5% above estimates at $826 million, as the company invested heavily on AI and its Rovo product. AI Race Sends Big Tech's Capital Spending to Stratospheric High "We've always been a business that is long on R&D," he said in the interview. "In the current AI era that we're in, it's incredibly dangerous to be spending a small amount on R&D if you're trying to adapt." When asked on an earnings call about market concerns about AI's impact on the software sector, Cannon-Brookes said all of the conversations with customers in the past year had been "incredibly positive," pointing to the increase in long-term customer commitments in the past year.
Share
Share
Copy Link
Atlassian CEO Mike Cannon-Brookes told investors the company can scale its Rovo agentic AI offering to five million users without inflating AI costs or hurting margins. Despite shares falling 70% over twelve months and $826 million in quarterly R&D spending, the CEO insists AI strengthens rather than threatens the collaborationware giant's business model.
Atlassian has demonstrated it can integrate AI into its services without sacrificing profit margins, a claim CEO Mike Cannon-Brookes emphasized during the company's second-quarter earnings call. The collaborationware company now serves five million users through its Atlassian Rovo agentic AI offering, yet has maintained stable gross margins despite the computational demands typically associated with AI deployment
1
. "We're able to deliver those five million Rovo seats and continue to improve gross margin," Cannon-Brookes told investors, crediting the company's engineering teams for this achievement1
.
Source: Bloomberg
The ability to manage AI costs effectively stems from operational optimization rather than simply passing expenses to cloud providers. Since Atlassian runs entirely in public clouds, this approach suggests the SaaS giant has found ways to add AI capabilities without dramatically increasing spending with major cloud operators
1
. This matters because it demonstrates that AI integration doesn't inevitably lead to margin compression, a concern that has plagued many software companies racing to adopt the technology.Despite maintaining cost discipline in deployment, Atlassian's investment in AI research and development remains substantial. R&D spending for the quarter reached $826 million, approximately 22.5% above analyst estimates, as the company poured resources into AI capabilities and its Rovo product
2
. Mike Cannon-Brookes on AI spending was unequivocal: "In the current AI era that we're in, it's incredibly dangerous to be spending a small amount on R&D if you're trying to adapt," he told Bloomberg Television2
.This aggressive R&D posture reflects Atlassian's conviction that AI represents an opportunity rather than a threat to its business model. The company's willingness to invest heavily while simultaneously controlling operational AI costs suggests a strategic bet on building proprietary capabilities that will differentiate its platform over time.
Cannon-Brookes used the earnings call and shareholder letter to push back against growing market concerns that AI could disrupt traditional SaaS businesses. "I'm convinced AI is great for Atlassian. Others think software is dead," he wrote in his shareholder letter, acknowledging that "noise swamps signal" in the current environment
1
. He argued that enterprise customers still value Atlassian's platform for its compliance, security, and trustworthiness—attributes that AI assistants alone cannot provide1
.The CEO also dismissed competitive threats from AI-native products like Anthropic's "Claude Cowork," which competes with Jira. Cannon-Brookes noted that such tools will still need to access data stored within Atlassian's systems, suggesting the company's position as a system of record provides inherent defensibility
1
. Customer conversations over the past year have been "incredibly positive," he said, with enterprises committing to longer-term contracts and adopting more of Atlassian's product stack2
.Related Stories
Atlassian reported Q2 revenue of $1.6 billion, representing 23% year-over-year growth, and reached the $1 billion milestone for cloud product revenue in a single quarter for the first time
1
. Operating losses narrowed from $57.5 million in Q2 2025 to $47.7 million1
. Cloud migrations remained healthy in the quarter as the company continues phasing out on-premises datacenter products1
.Yet tech shares have punished Atlassian severely. The stock has fallen more than 70% over twelve months and dropped an additional 8% in after-hours trading following the earnings announcement, bringing year-to-date declines to over 62%
1
2
. Cannon-Brookes expressed frustration with the valuation, calling the shares "significantly undervalued" and announcing accelerated stock buybacks1
.The disconnect between operational performance and market valuation reflects broader uncertainty about how AI will reshape the software industry. Investors appear concerned that heavy R&D spending and potential disruption from AI-native competitors could erode Atlassian's long-term position, even as the company demonstrates it can deploy AI at scale without destroying margins. Whether Atlassian's bet on platform durability and AI integration proves correct will depend on how quickly enterprises adopt AI-first workflows and whether Rovo can deliver enough value to justify continued investment.
Summarized by
Navi
[1]
31 Jan 2025•Business and Economy

11 Apr 2025•Technology

12 Oct 2024•Business and Economy

1
Technology

2
Policy and Regulation

3
Policy and Regulation
