Why AI agents love files
Why are the most advanced AI agents suddenly working with files, folders, scratchpads, and structured workspaces?
In this episode of AI Explainer, Ben Kus, BOX CTO, explores why modern AI systems are starting to look a lot more like human collaborators — navigating unstructured information, organizing tasks into folders, and producing outputs as documents, spreadsheets, and presentations.
From coding workflows to enterprise knowledge work, Ben breaks down why file-based systems have become such a natural fit for agentic AI, and the implications for the enterprise:
What happens when companies manage not just human employees, but thousands of AI agents working across sensitive HR, legal, and financial data?
The implications for security, permissions, workspace design, and multi-agent collaboration.
Why the future of AI may depend on structured environments that let agents work safely where data already lives.
Want to understand how AI agents will actually work inside real businesses? This interview lays a foundation.