This conversation features Mike Cannon-Brooks, co-founder and CEO of Atlassian, discussing AI transformation from dual perspectives: leading a company adopting AI internally and building AI infrastructure for 300,000+ organizations using Atlassian's tools like Jira, Confluence, Trello, Loom, and Rovo.

The discussion explores why context is becoming a core layer of enterprise AI, how agents and headless tool use are changing human-software relationships, the real constraints on enterprise AI adoption, and Mike's prediction that 2026 will mark AI's transition beyond chat interfaces into more natural product experiences.

Internal AI Adoption Challenges and Cultural Transformation

Enterprise AI deployment faces significant hurdles including talent changes, business process re-engineering, and security restrictions that initially limited Dia browser access to just 4 of 13,000 employees.

"We try to promote that loop at Atlassian... The learning about there's a new capability... you have to give them time and space and joy to play, to try it" - Mike emphasizes the importance of experimentation culture.

Successful AI adoption requires sharing both successes and failures across the organization since "we don't have people with 10 years worth of AI deployment experience inside the organization."

Meeting Enterprise Requirements While Driving Innovation

The Rovo platform must satisfy enterprise requirements including data residency, private model choice, customer managed keys, and compliance while enabling AI capabilities.

"Some of our customers take six months of queries to turn on our AI platforms. Some turn it on in a day" - highlighting the wide variation in enterprise adoption timelines.

Atlassian enables agents to be assigned to any JIRA work item, allowing existing workflows to be accelerated without requiring immediate process changes.

Teamwork Graph Evolution and Context as Competitive Advantage

"Acceleration for your business is about sort of intelligence multiplied by context" - Mike frames context as equally important to model capabilities for business value.

The teamwork graph now includes semantic indexing of entire codebases, complete organizational charts with skills mapping, and physical assets, operating at 150 billion object scale.

New CLI and MCP server capabilities allow the teamwork graph to be accessed by external agents and tools, enabling headless integration with coding agents like Cursor and Claude.

Characteristics of Leading AI Organizations

"The best customers ask us, How does Rovo connect to these 10 things?" - leading organizations think systematically about platform integration rather than tool proliferation.

Top-performing organizations focus on "20, 30% improvements in certain areas" rather than marginal 2% gains and emphasize output quality over token consumption.

DX acquisition enables understanding of engineering productivity metrics: "who's consuming the most tokens, which AI coding tools are working better" to measure actual impact.

The Future of AI Beyond Chat Interfaces

"We are going to start to see AI move beyond chat... bringing these technologies and powerful tools to everyday users in meeting them where they work" - Mike's key prediction for 2026.

Rovo Studio has enabled creation of over 5 million agents, demonstrating enterprise appetite for no-code AI solutions that don't require "crazy smart prompt engineer" skills.

The evolution includes disposable software concepts where "code is cheap, that can be thrown away and you can write another one tomorrow" through tools like Dia browser.