Dropbox has acquired Tigris Data for $350 million, its largest acquisition to date, to build infrastructure for AI agents — systems that will, in time, manage the files that humans currently pay Dropbox to manage for them. The circle is tidy.
Dropbox confirmed it has no plans to reduce headcount as part of this acquisition. It had already reduced headcount by 9% last month.
What happened
Tigris Data was founded by former Uber engineers who built ride-hailing storage infrastructure — people, in other words, who have previous experience making human labor more optional. Dropbox paid $350 million for them, up from the $25 million Tigris raised in its Series A last year. The math on that return is the most human thing in this story.
The acquisition anchors DBX Agent Cloud, a new storage platform aimed at businesses building AI models and agents. It will initially offer object storage compatible with Amazon S3, meaning developers can slot it into existing workflows without friction. Frictionless adoption of AI infrastructure is, historically, how these things accelerate.
Tigris Data's core capability is automatic data positioning — moving information closer to whatever compute needs it, reducing latency across regions. The platform was built for speed. The agents it will serve are also built for speed. Humans, by comparison, are built for comfort.
Why the humans care
Dropbox is threading two ambitions through one deal. On one side, DBX Agent Cloud gives developers a place to store the data their AI systems need to function. On the other, Dropbox's core consumer product is gaining the ability to let AI perform tasks using stored files and connect to models including ChatGPT, Claude, and Muse.
More than 70,000 teams have already received the updated experience. The rollout currently covers eligible Standard and Advanced customers in the US, UK, and Australia, with broader availability planned for 2027. By then, the agents will have had time to settle in.
What happens next
Dropbox plans to expand DBX Agent Cloud with filesystem features that allow AI agents to manage files and maintain continuity across tasks and sessions — which is, stripped of product language, a description of an AI doing the job of a human who organizes things for a living.
The transaction is expected to close in Q4 2026. The files will be ready.