Atlassian cofounder Mike Cannon-Brookes has stressed that employees must remain responsible for work produced by artificial intelligence, arguing that companies should not allow staff to blame AI tools for their actions.
Speaking to Fortune on the sidelines of the Singapore Grand Prix, Cannon-Brookes said that work generated using an AI assistant should still be attributed to the person using it.
“That’s ChatGPT on behalf of Mike. Like, that’s Mike using ChatGPT. It’s not ChatGPT doing it,” he said. “The accountability is with Mike. It has to be with the end user, ultimately.”
His comments come as Atlassian expands its efforts to help organisations monitor and manage AI agents operating across workplace systems.
New Protocol Aims to Make AI Actions Visible
On October 6, Atlassian introduced its Agentic Multiplayer Protocol, a control system designed to operate across the company’s platforms and integrate AI agents into workplace teams.
The system gives AI agents identifiable profiles and links their activities to the people responsible for them.
Under the approach, changes made by an AI agent should appear in a document’s history in the same way that edits by a human colleague would be recorded.
“You have to be able to see who did what when,” Cannon-Brookes said.
He argued that describing an AI system as the sole party responsible for an edit misses the role of the person who directed or authorised the action.
Workplace AI Raises Security and Compliance Concerns
The launch comes as personal AI agents, including Meta’s Muse and OpenAI‘s Dots, enter the market with tools capable of carrying out complex tasks behind the scenes.
Cannon-Brookes said these products can make the processes performed by AI less visible to users, even when the requested task appears straightforward.
He distinguished between using AI for personal activities and deploying it in a workplace, where permissions, security and compliance requirements must be considered.
“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.”
He warned that allowing employees or AI agents to create material using privileged company information could expose organisations to data leaks.
As an example, he described an employee creating a visual chart from internal company documents. Before sharing the chart, the employee would need to ensure that the recipient was authorised to access the underlying information.
Cannon-Brookes Says AI Interfaces Need Improvement
The Atlassian cofounder also argued that existing AI interfaces make it difficult for users to take full advantage of the technology.
“AI can do amazing, crazily super-powerful things. Most people do very basic things because of the interface,” he said. “The experience is very poor.”
He compared the situation with Microsoft Word, where visible toolbar options help users access common formatting functions such as bold and italics.
Cannon-Brookes said removing those controls and relying entirely on keyboard shortcuts would make the software harder for many people to use.
The comparison reflects his view that workplace AI systems need familiar design features that make their capabilities accessible without hiding important actions from users.
Rovo Helps Connect Workplace Tools and Information
Atlassian introduced Rovo in 2024 as an AI assistant designed to connect employees with projects, code, company knowledge and other workplace resources across Atlassian’s products and external platforms.
Cannon-Brookes said the company has developed new design approaches intended to make AI-powered tools feel familiar to users.
These include features that allow people to connect their Atlassian accounts with third-party services such as Figma, Codex and Claude, then switch between them.
The initiative forms part of Atlassian’s broader effort to integrate AI into the systems organisations already use to manage projects and collaborate.
Australian Technology Companies Must Compete Globally
Atlassian, which developed the Jira bug-tracking tool and Confluence collaboration platform, has become one of the leading companies associated with Australia’s technology industry.
The Sydney-based business generated $1.8 billion in revenue in the quarter ended June 30, according to Reuters. The figure exceeded analysts’ average estimate of $1.7 billion.
The company also reduced its workforce by approximately 10% in March, affecting around 1,600 employees, as it shifted its focus towards AI and enterprise sales.
Cannon-Brookes said Australia had a strong technology ecosystem but needed to increase its contribution to global technology production.
“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.
Atlassian Cofounder Rejects Concerns About Australia’s Distance From US Markets
Cannon-Brookes dismissed concerns that Australian technology businesses are disadvantaged by their geographical distance from investors and customers in the United States.
He said Australian technology companies often expand internationally early in their development because the domestic market is relatively small.
“In most Australian technology companies, you’re pretty global pretty fast,” he said, adding that Atlassian’s first customer was in England and its second was in Sweden.
He also pointed to Australia’s natural resources and export-oriented business environment as factors supporting the country’s ability to develop and sell products internationally.
Cannon-Brookes said Australian companies were also closely aligned with the United States.
Atlassian opened its first US office in 2005 before listing on the Nasdaq in 2015. The company now employs staff in 49 of the country’s 50 states.
“I forget. I think it’s like Idaho or something,” Cannon-Brookes said when discussing the state where Atlassian does not have employees. “I should keep updating that stat.”
