Dropbox brings AI into customer content and announces DBX Agent Cloud

AI actions inside the Dropbox workspace

The new Dropbox experience is designed to help customers find, understand, create, edit, organize, and act on content where their work already lives. Users can locate information, ask questions about files and projects, summarize folders, compare versions, and catch up on changes.

Dropbox also says customers can draft documents from project context, edit PDFs and images, transcribe recordings, and edit or refine video clips. They can ask Dropbox to build project folder structures, move or rename files, and carry out several steps from a single request.

These tasks use files and context already stored in Dropbox, so customers do not need to repeatedly upload and download content. Results remain connected to existing folders, permissions, and sharing context, allowing teams to review, collaborate, and continue working through their usual workflows.

More control over models and AI actions

Dropbox says the new experience has already rolled out to more than 70,000 teams. The initial rollout focuses on eligible Standard and Advanced teams in the US, UK, and Australia, starting in English. Broader availability is planned for 2027.

The company is adding controls intended to help customers understand and manage AI activity. Dropbox shows a plan for approval before making changes, works within existing permissions, lets customers undo changes, and provides a record of actions taken.

Customers can also choose and switch supported models within a session while keeping their work connected to its files, permissions, and collaboration context. Dropbox says supported AI tools including ChatGPT, Claude, and Muse can work with Dropbox files without requiring customers to re-upload them, with resulting work landing back in Dropbox.

DBX Agent Cloud targets AI application builders

DBX Agent Cloud extends Dropbox’s storage expertise to companies and developers building AI applications and agents. The offering starts with standalone object storage that customers can buy independently, and is designed to work alongside the clouds, compute providers, and developer platforms where applications already run.

The object storage service is S3-compatible, allowing developers to use familiar tools and supported applications built for Amazon S3. Dropbox says the service is intended to provide a consistent home for content across different environments, including training datasets, model artifacts, and multimodal content.

The planned acquisition of Tigris Data Inc. is intended to accelerate this expansion. Dropbox says it plans to add filesystem and content-processing capabilities next. Filesystem capabilities would give applications and agents a persistent place to manage files across sessions, while content processing would turn documents, images, audio, and video into useful inputs for AI.

Why the announcement matters

Dropbox is positioning the two initiatives around the same idea: keeping content, permissions, history, and context connected as AI tools and workflows change. For Dropbox customers, that means AI actions can take place within existing files and collaboration structures, with controls for approval, undoing changes, and reviewing activity.

For developers and companies building AI applications, DBX Agent Cloud offers a storage and data layer intended to operate across clouds and compute providers. Ashraf Alkarmi, co-CEO of Dropbox, said the company’s goal is to give customers “the freedom to choose how they use AI while keeping control of the content they depend on.”