Atlassian’s Mike Cannon-Brookes: AI agents need a name tag—and a human to be accountable for their mistakes

Atlassian cofounder Mike Cannon-Brookes is pushing back against the idea that artificial intelligence can become an office-level excuse when something goes wrong.

AI agents should not operate without ownership

Cannon-Brookes says AI agents need what he describes as a “name tag” along with a human who remains accountable for their mistakes. The message places responsibility for AI-related errors with people, rather than allowing employees or organizations to simply blame the technology.

His warning focuses on a basic challenge surrounding the use of AI agents: identifying who is responsible when their work produces an incorrect or unwanted result. Without clear ownership, “the AI did it” could become a way to avoid accountability in the workplace.

Why the message matters

For businesses using AI agents, the comments underline the importance of connecting these systems to identifiable human responsibility. The emphasis is not only on deploying AI, but also on making sure mistakes have an accountable owner.

Cannon-Brookes’ position offers a concise principle for workplace AI: technology may carry out tasks, but responsibility for its mistakes should remain attached to a person. The source does not provide further details about a specific Atlassian policy or implementation plan.