The SaaSpocalypse that wasn’t, with Atlassian CEO Mike Cannon-Brookes

Mike Cannon-Brookes is not buying the idea that AI will simply eat Atlassian. In a long chat with The Verge, the Atlassian cofounder and CEO argued that his company’s tools are built around something messier than raw task execution: the flow of work inside a business. Jira, Trello, and Atlassian’s other products are, in his telling, different ways of making that flow visible and manageable across teams, departments, and entire organizations.

That matters because the AI story around software has gotten weirdly apocalyptic. If a model can read across a company’s systems, summarize the important bits, and maybe even spin up new tools on demand, then why keep paying for a platform in the middle? Cannon-Brookes’ answer is that most businesses are not simple enough for that fantasy. They’re full of rules, compliance, exceptions, and human judgment. AI can handle the routine parts, but the hard parts still need people who know what the business is actually trying to do.

He kept coming back to the same idea: Atlassian’s job is legibility. The software is a “human reference to work,” as he put it, which means it helps managers and teams understand what’s happening at different levels of the business. AI may take over pieces of a process — Cannon-Brookes used the example of a sales deal exception workflow — but someone still has to know how many exceptions were granted, where the bottlenecks are, and whether the process itself needs to change.

That leads to his bigger claim, which is less dramatic than the SaaSpocalypse crowd would like and probably more plausible: AI will increase usage of Atlassian’s tools, not erase them. Faster processes, better quality, more consistency, more output. But also more need for humans to make judgment calls. And because he clearly enjoys a good argument, Cannon-Brookes pushed back on the idea that frontier models are about to run whole businesses by themselves. If that were true, he said, the businesses would have to be fairly simple. Most aren’t.

He also linked the AI shift to a familiar tech pattern. Mobile banking didn’t kill branches; it changed what branches were for. Same with enterprise software, he argued. The old jobs may shrink, but the systems don’t vanish. They adapt, and the work moves up a level.