Atlassian's Oct. 6 Protocol Makes AI Agents Show Their Name—and Their Owner

By Amara Osei

Atlassian shipped its "Agentic Multiplayer Protocol" on Oct. 6, a control plane that turns AI agents into visible, governed team members with distinct profiles and owner attribution.

The cofounder's point is simple: no more "the AI did it." Mike Cannon-Brookes told Fortune, on the sidelines of the Singapore Grand Prix: "That's ChatGPT on behalf of Mike. Like, that's Mike using ChatGPT. It's not ChatGPT doing it."

"The accountability is with Mike. It has to be with the end user, ultimately," he said.

In practice, an agent's edits should appear in a document's history the way a colleague's edits would. "You have to be able to see who did what when," Cannon-Brookes said. When software reports that Codex made an edit, his answer is blunt: "No, it didn't—someone did that."

Why does attribution matter now?

A wave of personal AI agents is hitting the market—Meta's Muse and OpenAI's Dots among them. Cannon-Brookes sees a design flaw in those products: they're "taking an agentic harness and change the interact to make it invisible, as to what it's doing." Agents run complex operations behind the scenes even when the user's request is simple.

That's tolerable for consumers. It's not for the workplace. "If you're making something on Muse, it's probably just for you," he said. "In a workplace, there's a lot of conversation around permissions, security, and compliance."

The risk is concrete. An employee—or an unsupervised agent—building content from privileged company data can leak it fast. Cannon-Brookes gave one example: a worker creating a visual chart from internal documents. "But if that's based on [internal] documents, you've got to be very careful if you're sending that to anybody, they can see the same documents, and you don't want to leak information," he said.

There's also an interface problem. "AI can do amazing, crazily super-powerful things. Most people do very basic things because of the interface," Cannon-Brookes said. "The experience is very poor."

His analogy is the word processor. "If you think about Microsoft Word with the toolbar, you learn how to use bold and italics," he explained. "If all that disappeared and it was only keyboard shortcuts, everyone's documents end up looking like plain text."

Atlassian launched Rovo, its in-house AI assistant, in 2024. Rovo connects people, projects, code and company knowledge across Atlassian and external platforms. "We've had to invent a lot of new foundational design patterns that feel familiar," Cannon-Brookes said—citing features that link Atlassian accounts to third-party tools like Figma, Codex or Claude, with seamless toggling between them.

What's the business backdrop?

Atlassian built its name on Jira, the bug-tracking tool, and Confluence, the collaboration platform. The Sydney-based company pulled in $1.8 billion in revenue for the quarter ended June 30, beating analysts' average estimate of $1.7 billion, according to Reuters.

The quarter came after a hard pivot. Atlassian cut its workforce by 10% in March—roughly 1,600 employees—as part of a shift toward AI and enterprise sales.

Cannon-Brookes frames Atlassian as the standard-bearer for an Australian tech scene that now includes design platform Canva and payments firm Airwallex.

"Australia has an amazing technology ecosystem," he said. "We're not going to build it all since we're quite a small country… but we're good at taking operating systems, databases and AI models, and applying them to problems to create economic value and GDP."

"We're roughly 1% of the world's GDP. We need to be 1% of the world's technology production," he added.

He dismisses the idea that distance cuts Australian firms off from U.S. investors and customers. "In most Australian technology companies, you're pretty global pretty fast," he said. "Our first customer was in England, and our second customer was in Sweden… We don't have much of a consumptive market in Australia so we're big exporters and travelers, but also big makers of things since we have lots of great natural resources."

Atlassian opened its first U.S. office in 2005 and listed on the NASDAQ in 2015. Today it employs people in 49 of the 50 U.S. states. "I forget. I think it's like Idaho or something," Cannon-Brookes said of the lone holdout. "I should keep updating that stat."

The bet behind the Oct. 6 protocol is that enterprises will only let agents into core workflows once every action carries a name tag—and a human to answer for it.