For nearly two decades, over 700 million registered users have trusted Dropbox to store, organize, and share their most important work. AI is fundamentally changing what’s possible with that content, and Dropbox is evolving to meet that moment.

Today, Dropbox is introducing a major evolution, bringing AI-powered action directly to users’ existing files and context to help them move beyond finding and understanding information to getting work done.

Users can now tell Dropbox what they want to accomplish, not just what they want to find. Dropbox identifies relevant content, proposes a plan, and, with their approval, takes action directly on existing files. Content stays in Dropbox, with existing permissions, versions, and context intact. Users can review the actions taken and undo supported changes.

By connecting files, business applications, and AI tools, Dropbox is becoming a trusted content and action layer that helps users turn the content they already have into meaningful outcomes.

“We’re bringing AI directly to the content and context you already have in Dropbox, helping you not only find information, but get to a final outcome while keeping full control over the steps it takes to get there. We want you to have the freedom to choose how you use AI, with the peace of mind that your content stays connected, protected, and under your control,” said Ashraf Alkarmi, co-CEO of Dropbox.

AI that acts, with users in control

Giving AI the ability to take action on business files requires trust. Dropbox works within existing permissions, shows users a plan to approve before making changes, and maintains a record of the actions taken. Users can review changes through version history and undo supported actions when needed.

These safeguards provide transparency and control while keeping files, permissions, and collaboration intact.

At creative agency YYES, Co-founder Brent Stickels saw the benefits firsthand. Dropbox reorganized a project folder of approximately 100 files in two or three minutes, compared with the 15 to 20 minutes it took Brent to sort the same folder manually. Just as important, he remained in control of the changes.

“Before it changed anything, the AI in Dropbox told me its plan and asked me to approve it. Kicking the tires like that gave me faith that it’s really smart and that it checks in with me to make sure it’s doing the right stuff,” said Brent Stickels, co-founder of YYES. 

Your files stay in Dropbox, your permissions remain in place, and changes stay synced across devices and shared folders.

Content stays connected. AI choice stays flexible. 

As AI tools and models continue to evolve, users shouldn't have to move their content every time they change tools. Dropbox provides a consistent home for files, permissions, version history, and collaboration while giving users the flexibility to choose how and where they work with AI.

Within Dropbox, users can choose and switch between supported AI models in the same session without losing the context of their work. They can use these models directly in Dropbox without a separate account or paid subscription from the AI provider. Dropbox also connects with business applications like Gmail, Google Drive, Outlook, Teams, and Slack, bringing relevant information into workflows when needed. 

Dropbox meets users in the AI tools they already use. Through supported integrations with ChatGPT, Claude, Gemini, Muse, Cursor and Copilot, users can work with their Dropbox content without repeatedly downloading and uploading files.

For Let’s Cook Video Editing Services, connecting Dropbox with Claude is helping streamline project setup. Instead of manually creating folders for each new project, the team can use AI to get its Dropbox workspace organized in a fraction of the time.

“A project setup task that used to take us around 15 minutes can now be done in about a minute with Claude and Dropbox. We can create the Dropbox folder system for a new project in a matter of seconds,” said Aleksandra Lason, Chief Operating Officer at Let’s Cook Video Editing Services.

Availability

Dropbox’s new AI-powered capabilities have already rolled out to more than 70,000 teams. The initial rollout focuses on businesses on Standard and Advanced plans in the US, UK, and Australia, starting in English, with broader availability planned for 2027.

More information about the new Dropbox capabilities and availability is available at http://dropbox.com/a-smarter-dropbox.

Dropbox now brings AI directly to customers’ existing content, with greater choice, control, and peace of mind. DBX Agent Cloud extends our infrastructure expertise to companies and developers, starting with object storage.

Today, we’re taking two steps to make that possible:

• We’re rolling out the new Dropbox experience, helping customers find, understand, create, and edit content directly where their work already lives, with model choice and safeguards that give them confidence as AI takes action

• We’re announcing DBX Agent Cloud and the planned acquisition of Tigris Data Inc., extending Dropbox’s infrastructure expertise to companies and developers, starting with standalone object storage and adding filesystem and content-processing capabilities next

AI tools, models, and compute environments will continue to change. The content those systems depend on, along with its permissions, history, and context, needs a consistent home.

“The value of AI depends on the content and context it can work with,” said Ashraf Alkarmi, co-CEO of Dropbox. “We’re bringing AI to the work people already keep in Dropbox and extending our storage expertise to companies building AI applications. Across both, our goal is to give customers the freedom to choose how they use AI while keeping control of the content they depend on.”

AI takes action directly on existing Dropbox content

Dropbox serves customers whose work depends on content, including large files, rich media, specialized file types, and collaboration with external partners. Dropbox brings AI directly into that work. Now, customers can prompt Dropbox to:

• Find and understand: locate information, ask questions about files and projects, summarize folders, compare versions, and catch up on changes

• Create and edit: draft documents from project context, edit PDFs and images, transcribe recordings, and edit or refine video clips

• Organize and act: build project folder structures, move or rename files, and carry out several steps from a single request

Customers can complete these tasks using the files and context already in Dropbox, removing the need to repeatedly upload and download content. Results stay connected to the same folders, permissions, and sharing context, so teams can review, collaborate, and continue their work.

The new Dropbox 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, with broader availability planned for 2027.

Model choice and controls over AI actions

As AI takes on more work, customers need clear ways to understand and manage what it does.

Dropbox shows a plan for approval before making changes, works within existing permissions, lets customers undo changes, and provides a record of the actions taken. These safeguards help customers use AI with confidence and help organizations keep content protected and governed.

Customers also have greater freedom to choose how they use AI. In Dropbox, they can choose and switch supported models within a session while keeping their work connected to its existing files, permissions, and collaboration context.

Dropbox connects content to supported AI tools

Customers can also work on their Dropbox files in supported AI tools, including ChatGPT, Claude, and Muse, among others, without re-uploading their content, with resulting work landing back in Dropbox.

Teams can then share, review, refine, and reuse that work through their existing Dropbox workflows. DBX Agent Cloud extends that principle of choice to the infrastructure companies use to build applications.

DBX Agent Cloud: data infrastructure for AI applications and agents

AI applications and agents create new demands for storage and data access. Training datasets, model artifacts, and multimodal content need to work across clouds and compute providers, while agents need persistent access to files across tasks and sessions.

DBX Agent Cloud brings Dropbox’s experience operating storage infrastructure at scale to companies and developers, starting with standalone object storage they can buy independently. The planned acquisition of Tigris would accelerate this expansion and advance our vision to be the trusted, independent content foundation for the AI era. The object storage offering is S3-compatible, allowing developers to use familiar tools and supported applications built for Amazon S3.

DBX Agent Cloud is a storage and data layer, designed to work alongside the clouds, compute providers, and developer platforms where customers already build and run their applications. Our focus is on giving those applications a consistent home for their content, wherever they run.

“AI is driving storage demand at a scale this industry has not seen before. Meeting it takes massive scale and a conviction we share with Dropbox: that customers should have the flexibility to run their compute wherever it serves them best, with a product built for developers and agents alike,” said Ovais Tariq, co-founder and CEO of Tigris. “Dropbox brings nearly two decades of operating storage at that scale. We bring the developer-focused product our customers depend on. Together we would meet that demand across the industry, and deliver the capabilities that come next, starting with filesystem and content processing.”

We plan to add filesystem and content-processing capabilities next. Content processing helps turn documents, images, audio, and video into useful inputs for AI. Filesystem capabilities give applications and agents a persistent place to manage files and continue work across sessions.

With these updates, we’re expanding how people and companies put their content to work by bringing AI directly into Dropbox, connecting Dropbox files to supported AI tools, and extending our infrastructure expertise to companies building applications. Across these efforts, we’re helping customers keep their content connected and under their control as AI tools, models, and workflows evolve.

Today, co-CEO Ashraf Alkarmi shared the difficult news with employees that we’re reducing our global workforce. His email is below.

Hi everyone,

Today, we’re making difficult choices about where we invest and how we organize Dropbox to achieve the product and growth priorities that will matter most for our long-term strategy. As part of these changes, we’ve made the decision to eliminate about 200 roles across Engineering, Product, Design, and G&A teams (approximately 9% of our global workforce). For most impacted Dropboxers, today will be their last day.

S-team and I take full ownership of this decision and recognize that these colleagues have made meaningful contributions to Dropbox, our products, our customers, and our teams. Our first priority today is to treat every person affected with the care and respect they deserve.

Changes to Senior Leadership

I also want to share that Ali Dasdan has decided to move on from Dropbox. During his time here, Ali led and strengthened many of the technical capabilities we’ll rely on as we move into the next phase of our transformation. I’m grateful for his leadership, his partnership, and the foundation he and the team have put in place. Ali’s last day is November 18th, and he’ll be focused on ensuring a smooth transition through next month. We wish him all the best in what comes next.

Why we’re making these changes

Our business is profitable, and we’ve built real momentum this year. Our two-year strategy sets a clear direction for growth: solving more of our customers’ critical workflows; reaching more of our chosen customers, including larger companies; and bringing our technology to new markets.

Delivering on that strategy requires concentrating our resources on the priorities that will have the greatest potential for customer impact and growth.

That means making hard choices about where we invest, how we’re organized, and what work we simplify, scale back, or stop. Those choices include the difficult decision to reduce the size of our team.

How we’ll prioritize going forward

In order to be successful, we can’t take the same work and spread it across fewer people. We’ll streamline teams and concentrate support and resources against product areas that have the highest impact on our customers and Dropbox as a business.

Taking care of impacted employees

To those leaving Dropbox, we're committed to supporting you through this transition. Departing Dropboxers are eligible for a severance package in line with the following:

• Severance, equity, and transition support:
  • Sixteen weeks of pay beginning today, plus one additional week for each completed year at Dropbox. International packages will reflect local practices and legal requirements
  • Q4 equity vesting
  • For employees on the Corporate Bonus plan, a prorated lump-sum payment based on their 2026 bonus target
  • Immigration support for impacted visa holders
• Healthcare and well-being:
  • U.S. employees will receive six months of COBRA subsidy, if elected and eligible
  • Modern Health access will continue through the end of the separation month to support mental well-being, with extension options detailed in severance packages
  • Payment for eligible current and approved upcoming paid leaves, including medical and family leave
• Additional support:
  • Devices: Impacted employees will be eligible to keep (1) Dropbox laptop, (1) mobile device if assigned, and all existing desktop peripherals
  • Job placement: Job placement services and career coaching will be available at no cost

What comes next

Right now, our focus is on the colleagues who are leaving Dropbox. Please give them the space and support they need.

And to everyone leaving Dropbox: thank you. Your work to build our capabilities, serve our customers, and contribute to our culture has been deeply meaningful. I’m grateful for everything you’ve contributed.

Ashraf

Important work rarely happens in a single app. Your campaign briefs, project specs, and historical decisions live in Dropbox, while your team creates presentations, spreadsheets, and briefings in Microsoft 365. For many of our customers, Dropbox and Microsoft 365 already work side by side. Strengthening that connection across the workflows they rely on has been a clear priority.

Today, we’re introducing Dropbox for Microsoft Copilot, a new integration that lets teams bring trusted Dropbox context into Copilot, so they can create from information their business already relies on.

Users can now ask Copilot to find, open, summarize, organize, and compare Dropbox content, then use that project context to create new work across Microsoft 365 applications. They can also save supported text-based outputs back to Dropbox and create shared links for collaborators.

The value goes beyond easier file access. Teams can use the context already in Dropbox to create new work in Copilot without separating that work from the sources, permissions, and controls their business relies on. Dropbox remains a durable, governed home for the source material and supported outputs, ready for review, sharing, and the next task.

Move from trusted context to completed work

For businesses, finding the right information is only the starting point. The real value comes when teams can turn trusted project context into work they can review, share, and use again.

The Dropbox integration for Copilot connects the project knowledge in Dropbox to the Microsoft 365 tools teams use to create. Users can draw on relevant context to develop presentations, analyses, and briefings, while supported text-based outputs, including CSV files, can be saved back to Dropbox as part of the project record. This continuity reduces manual handoffs, repeated context assembly, version confusion, and avoidable rework.

For example, a marketing lead can reference an approved campaign brief, performance data, and creative assets in Dropbox to identify key insights and create a Microsoft PowerPoint presentation grounded in the project’s direction.

An architecture or construction team can reference project specifications and recent updates in Dropbox to identify changed requirements, prepare a stakeholder briefing, or generate a CSV for analysis in Microsoft Excel. The CSV can then return to Dropbox for review and sharing.

The result is a more connected workflow. Copilot can help teams act on the context in Dropbox, while Dropbox keeps the underlying sources and supported outputs connected for what comes next.

"Teams work across multiple applications, and AI should help bring relevant context directly into the flow of work," said Chantrelle Nielsen, Principal Product Manager, Copilot Platform Microsoft. "Dropbox for Microsoft Copilot helps teams seamlessly connect their content to create presentations, develop analyses, and make informed decisions across Microsoft 365, while maintaining established permissions and access controls. Together, we’re helping organizations bridge context with action and keep work moving."

Keep project context connected and access governed

For AI to be useful at work, organizations need confidence that connecting their tools will not change who can access the underlying information. The Dropbox integration brings project context into Copilot while keeping access tied to the permissions already in place.

After users connect and authorize their Dropbox account, Copilot can only work with files and folders they are authorized to access. Existing Dropbox permissions, sharing settings, team policies, and admin controls continue to apply. Connecting the integration does not give users broader access or change access to the source files. That matters when projects span departments, client teams, and external partners. People can use the Dropbox context relevant to their work without making it available to everyone using Copilot.

As organizations adopt different AI tools for different tasks, Dropbox provides a consistent, customer-owned home for source material and supported outputs. This helps businesses keep their work connected and governed without rebuilding their project context or access controls in every new system.

Create from the Dropbox context your business trusts

Work may move across tools, but its context should stay connected. The Dropbox integration for Copilot lets teams use trusted project context across their workflows and save supported outputs back to Dropbox, keeping source material, permissions, and new work connected in a customer-owned home.

The integration is available now for customers on all Dropbox plans. Appropriate Microsoft Copilot license and additional access may be required to use in Copilot Cowork. Visit our Help Center to learn how to connect Dropbox to Microsoft Copilot.

Dropbox connects trusted project context to a growing range of AI tools, including Microsoft 365 Copilot, ChatGPT, Claude, Meta Muse, Google Gemini, and more. This gives teams the flexibility to work with the tools that fit their needs while Dropbox remains the foundation for their projects. Explore Dropbox AI integrations.

At Dropbox, we’ve spent a lot of time learning what people need to do their best work in a Virtual First world. As the landscape continues to evolve, one question keeps us curious: What does it take to thrive in the AI era? Here, we take a look at how managers help create the conditions for success. We explore the traits and habits of those who lead our strongest contributors, and what you can learn from them to support your own team.

Recap: What makes people thrive at work?

Last year, we looked at a group we call thriving employees—Dropboxers who show both high engagement and high performance. We wanted to understand what sets them apart and the role AI plays in their daily work.

Our research showed that thriving wasn’t simply about using AI more often. The difference showed up in how these employees worked. They used AI to elevate their impact, but they also protected time for focus and recovery, invested in relationships and growth, and leaned into human capabilities like awareness, adaptability, judgment, and connection. We call these durable skills, and they become even more valuable as technology evolves.

New research: Understanding the manager effect

This year, we looked at the people who manage these successful Dropboxers. Our latest internal research suggests they share many of the same durable skills.

Managers of thriving employees:

• Were 85% more likely to be thriving themselves than other managers
• Were more than twice as likely to invest AI time savings in coaching, one-on-ones, and development
• Had a clearer view of team performance, with 92% of our most senior managers reporting clarity on performance against expectations, compared with 74% of other senior managers

AI is accelerating the creation of business content and raising the stakes for how that content is found, shared, and put to work. Every proposal, analysis, and project plan draws on information a business already holds and creates new knowledge for others to use. As content volumes grow, time spent gathering context, locating the right version, and moving useful output between tools can add up across teams. Our research explores how AI is changing content at work, where that friction is building, and what it means for businesses aiming to turn faster creation into lasting productivity.

Dropbox surveyed 219 professionals primarily working in technology, AI, software, finance, and healthcare at The AI Conference 2026 in San Francisco. This report brings together 70+ statistics from our latest market research on how AI is changing content at work—from time spent gathering context to the growth in content creation and where useful AI-generated work is saved and shared.

Key report findings

• 15.8 hours per week averaged engaging with AI at work
• 36.6% of AI engagement time is spent unproductively retrieving context
• 4.3 additional context inputs required after the first prompt
• 46.4% of the new work files are developed with AI
• 162.6% average growth in new work files created since AI adoption
• 93.2% of people create more work files now with AI
• 43.8% of useful results created by AI never leave the prompt

AI is becoming a substantial part of the workweek, and the volume of content people create is rising with it. Gathering context is a major source of friction, with people repeatedly adding files, links, and background before a result is ready to use. Useful output then spreads across shared workspaces, local files, and chats, leaving teams with a growing body of knowledge to organize, share, and put to work.

We spend 15.8 hours per week at work actively engaging with AI

What does your average workweek look like now compared with three years ago? While many professionals previously lost many hours to routine jobs reviewing, writing, researching, or developing actual code, a lot of this work is delegated to AI these days. But with AI optimizing so many areas of work, how much time do we have to invest in actively managing our AI? 

The average person spends 15.8 hours a week actively engaging with AI at work alone—and this is outside of auto-running Codex jobs that operate on autopilot. But even more telling, only 2.3% of people report zero time with AI on an average week.

Here’s what people report for AI engagement time on an average week:

• 2.3% spend 0 hours at work engaging AI
• 16.4% spend 1–5 hours at work engaging AI
• 24.7% spend 6–10 hours at work engaging AI
• 7.8% spend 11–15 hours at work engaging AI
• 16.4% spend 16–20 hours at work engaging AI
• 6.4% spend 21–25 hours at work engaging AI
• 5.0%  spend 26–30 hours at work engaging AI
• 21.0% spend 30 or more hours at work engaging AI

When you work for yourself, the work behind the work adds up fast. Finding the latest brief. Chasing missing files. Organizing project materials. Getting everything ready to share. AI agents can help carry that work forward, but they’re most useful when they have the right context: the briefs, drafts, references, and project history you already keep in Dropbox.

Today, we’re launching the Dropbox connector for Muse, giving you a new way to put the files and project history you already keep in Dropbox to work.

Connect Dropbox to Muse, and your personal agent can use your project context to find relevant files, organize project materials, collect what’s missing, prepare work to share, and carry out actions as part of a broader, multi-step plan.

The idea is simple: start with the work you already have instead of rebuilding the background from scratch.

"For a small business, the context behind the work already lives in Dropbox, from creative assets to client files and agreements. Connecting Dropbox to Muse brings that trusted context into another place customers choose to work, helping them find, create, and organize what they need without jumping between tools. The opportunity is to make Dropbox useful wherever work happens, so customers can spend less time managing it and more time growing their business." — Ashraf Alkarmi, co-CEO, Dropbox

Give Muse the goal. Start with the work already in Dropbox.

Say you run a small business and you’re onboarding a new customer. Your onboarding checklist, welcome materials, service information, and project templates are already in Dropbox. Instead of managing every step separately, you could ask Muse: "Onboard this customer using our materials in Dropbox."

With the Dropbox connector enabled, Muse can use those materials to identify the required steps, information, and project milestones. It can create a customer folder, organize the relevant documents, set up a file request for anything that’s missing, and prepare a shared link for the customer. Muse can also draft a welcome note and kickoff checklist or coordinate next steps with other connected apps, where supported.

Dropbox gives Muse more than the document that starts the process. It provides the source materials and project history behind the work, connecting what was agreed to incoming assets, new work, and final deliverables as the project moves forward. That means Muse doesn’t have to start with an isolated prompt or ask you to reconstruct the background every time.

Muse can help plan and coordinate the steps using the context available through Dropbox. For supported workflows, Muse is designed to continue working after you close the app and return when it needs your input or approval.

Build on the work you've already created

The same applies to creative work. A long-form video in Dropbox might sit alongside brand guidelines, past scripts, campaign plans, and reference assets. Ask Muse to "turn this into a week of content," and it can work from that existing body of work instead of starting from a blank prompt.

Muse can identify key themes and help prepare a video description, newsletter copy, social captions, and short-form concepts. Through the Dropbox connector, it can organize those materials into a campaign folder alongside the original recording, keeping the source, brand context, and new work together.

For freelancers and small-business owners managing several projects at once, Muse can also help create a clearer view of what needs attention. Ask it to surface missing client inputs, upcoming deliverables, and decisions that need your attention using the project materials in Dropbox. You spend less time reconstructing where each project stands and more time acting on what needs to happen next.

AI models and agents are only as smart as the context you give them. The latest file rarely captures the full project history stored in Dropbox: changing briefs, shifting feedback, tradeoffs, and approved decisions. Without that history, AI-assisted work can move backward by reviving rejected messaging, missing a change order, or following a ticket without understanding the requirement behind it.

We’re collaborating with SpaceXAI to introduce two new integrations to close that gap: Dropbox in the Cursor Marketplace and the Dropbox MCP for Grok Bot. 

AI-assisted work should draw on the project history people already store in Dropbox, not just the last file they worked with. As the trusted context layer for AI, Dropbox connects that history to the tools people choose.

With Dropbox in the Cursor Marketplace, technical teams can use requirements, research, and decisions from Dropbox as they build. The Dropbox MCP for Grok Bot lets people give Grok Bot a role and a goal without repeatedly finding files or re-explaining the project. The underlying source material stays organized in Dropbox and existing permissions continue to apply.

Results can return to Dropbox, where they stay connected to what came before and are ready to review, share, and use again.

Build from the context that shaped the requirements

A codebase can show how a product works, but not always why a requirement changed or which tradeoff was approved. That knowledge often lives in product briefs, API documentation, research, meeting notes, and client requirements stored in Dropbox. Without it, technical teams can lose time reconstructing decisions or risk building from an incomplete picture.

Dropbox in the Cursor Marketplace brings the materials around the code into AI-assisted build workflows. A developer can work from requirements and technical-review notes while implementing a feature. A product manager can draw on briefs, customer research, and technical decisions while drafting a specification or preparing a prototype.

They can also organize supporting materials and save new documentation back to Dropbox. When tradeoffs are made, they can record the reasoning in a README, decision log, project note, or updated specification. The integration includes 19 tools and six workflow-based skills for work like this.

Saving that reasoning to Dropbox turns it into durable documentation that colleagues, reviewers, and future contributors can use to understand how the work took shape.

“Building great software takes more than code. Teams also need the decisions and project history behind the work,” said Toni Adams, Head of Global Partnerships at SpaceXAI. “Dropbox in the Cursor Marketplace makes it easier to bring that context into the development process, capture new decisions, and keep the work connected for everyone contributing to the project.”

From goal to action, with context already in reach

An AI assistant needs more than a goal; it needs the background behind the assignment. Campaign planners and project managers rarely work from a single file; they rely on briefs, assets, research, meeting notes, feedback, and decisions gathered over time.

The Dropbox MCP for Grok Bot makes the materials and decisions already organized in Dropbox available as the task unfolds. That lets users move directly into the assignment instead of spending the first few prompts assembling background.

A marketing lead could ask Grok Bot to develop a plan informed by approved messaging, current assets, and earlier campaign research. A construction lead could ask it to assemble an owner briefing from specifications, meeting notes, and change documentation. In each case, Grok Bot can move the work forward using relevant project history from Dropbox, not just what’s included in a single request. 

The result can then be saved back to Dropbox for review, sharing, and next steps instead of remaining isolated in a Bot response.

Keep access grounded in Dropbox permissions

Your existing Dropbox permissions apply to both integrations. They can only access content the signed-in user can already access in Dropbox, and any access policies set by their organization remain in effect. Users must also authorize the connection and approve the required scopes. Any deletion must be reviewed and explicitly confirmed by the user. 

As AI moves from answering questions to taking action, these safeguards keep consequential actions visible and in the user’s hands.

Getting started

Dropbox in the Cursor Marketplace
Dropbox in the Cursor Marketplace is available to individuals and organizations on all Dropbox and Cursor plans, with no current country restrictions.

Dropbox MCP for Grok Bot
The Dropbox MCP for Grok Bot is available across eligible Cursor individual, team, and Enterprise subscriptions: Cursor Pro, Pro+, and Ultra; Cursor Teams Standard and Premium; and Enterprise.

Connect Dropbox in the Cursor Marketplace to access the Dropbox MCP across both apps.

Learn more about the Dropbox and SpaceXAI partnership.

AI is supposed to reduce busywork, but people are still doing much of the manual work required after AI creates an output. A Dropbox-sponsored study found that while 54% of professionals now start knowledge work with AI, 74% manually move AI output between apps at least three times on an average AI workday, and only 11% say AI produces finished work without further barriers. 

A Dropbox-sponsored study of more than 500 US professionals who use AI reveals just how wide the gap is between where work starts and what it takes to finish it.

AI can generate an answer in seconds, but getting that output across the “last mile” into finished work still requires the right context, verification, people, permissions, and workflows. Today, that often means moving content between apps, sharing it with team members, and getting it ready for review. 

What we found

• AI is becoming where work starts, but rarely where it finishes. More than half of respondents (54%) say an AI assistant is now their primary starting point for new knowledge work, ahead of documents, email, search engines, and project management tools. Yet only 11% say AI already generates finished work without further barriers. 

• The last mile remains manual. Nearly three-quarters (74%) manually move AI-generated work between applications at least three times on an average AI workday. For 30%, that happens six or more times, while 65% use three or more applications to complete a typical AI-assisted task.

• AI output is becoming organizational knowledge. Forty-three percent say most of their weekly AI-assisted work becomes permanent team documentation or part of a shared knowledge base.

The new "last mile" of work

That work after generation is AI’s new “last mile.” Output still has to move through the context, verification, people, decisions, permissions, and workflows required to make it trusted and usable. As AI tools expand beyond generation, the challenge is keeping that work connected across the tools, systems, and people teams rely on. That’s where Dropbox comes in, connecting people and content so work can keep moving. 

Consider a customer proposal. AI might generate the first draft, but someone still needs to check it against customer information, gather feedback, incorporate revisions, get approval, and save the finished version somewhere the team can find and reuse.

We’re building Dropbox to play a broader role in that last mile by helping content and context stay connected across AI experiences, while giving teams a consistent place to continue, share, and build on their work.  

AI has already changed the first mile of work. The next opportunity is making the last mile as seamless as the first.

~ ~ ~

About the research

Dropbox sponsored a survey of 504 AI-using US B2B professionals from July 28 to July 30, 2026. Respondents represented functions including IT, engineering, sales, marketing, finance, operations, HR, and customer support.

AI is only as smart as the context you give it. Too often, you’re stuck copy-pasting brand guidelines, searching through old folders, or settling for generic drafts that ignore your past work. 

The result is work that's scattered and disconnected from the source material and project history you rely on every day. 

That’s why we’re launching new Dropbox integrations for Google Gemini. You can connect Dropbox to Gemini to bring the context you already keep in Dropbox into your work, so you can find, summarize, and create from your own projects and past work without leaving the conversation. 

We're bringing Dropbox into Gemini through two experiences: Gemini chat and Gemini Spark.

The Dropbox app for Gemini is available to anyone with a personal Google Account who is 18 or older in the US, making it easy to reference your Dropbox content while working in Gemini. For Google AI Pro and Ultra subscribers, Gemini Spark takes it further, moving projects forward as part of automated, multi-step workflows.

Together, these experiences let you bring your Dropbox files, source material, and project history into Gemini and turn that context into your next piece of work, while keeping Dropbox as the trusted home for everything behind it.

Create from your own content with Gemini

Gemini helps you use your Dropbox context to research, plan, and draft, so you can spend less time hunting through folders and more time creating.

If you’re a content creator, for example, your Dropbox already holds the working history behind your creative projects: scripts, brand guidelines, sponsor briefs, past videos, research, and photos. Instead of digging for the right file, you can ask Gemini chat to pull it up (e.g., “find the latest version of my channel style guide”), or to summarize a long sponsor brief before you start filming. Because Gemini is working from your actual files, your output is instantly grounded in your personal context, not a blank slate.

You can also turn what you’ve already made into what’s next. Ask Gemini to analyze past video scripts stored in Dropbox and pitch new angles, or draft social captions that reflect your brand voice. Rather than starting from scratch, you start from your own catalog of work, and when a draft is ready to share, you can export it to a Google Doc in a click.

Automate multi-step workflows with Gemini Spark

Some work takes more than finding information. It takes organizing all the material behind a project, prepping it to share, and keeping the project moving. Gemini Spark helps you go beyond the conversation, working with Dropbox to organize files, generate shareable links, create file requests, and keep your work on track.

Imagine you’re a freelance designer kicking off a new project. You’ve got a brief to review, reference images and past work to pull together, and new assets still coming in. Rather than sorting it all out by hand, you can ask Gemini Spark to find and summarize the brief, create a project folder, organize your files by stage, set up a file request for missing assets, and generate a shareable link for the final deliverables.

Because Spark works with Dropbox, Dropbox stays the trusted source for your files and project work, helping keep everything from getting scattered across disconnected AI tools. Gemini’s access stays aligned with the files and folders you’re permitted to access in Dropbox. And before taking sensitive actions, like sharing, deleting, or changing permissions, Gemini asks you to confirm before it acts, so you’re always in control.

Getting started

The Dropbox app for Gemini is available to anyone with a personal Google Account who is 18 or older in the US. Gemini Spark is available to Google AI Pro or Ultra subscribers in the US. For this initial release, the Dropbox connected app is available in English only, in the Gemini web app at gemini.google.com, and requires a personal Google Account with Keep Activity turned on. 

With Dropbox connected to Google Gemini, you can put the context behind your work to use while Dropbox keeps it all together in one trusted home.

Get started by connecting Dropbox to Gemini. In Gemini, go to Settings → Personal Intelligence → Connected Apps → Dropbox to connect. 

Learn more about our partnership with Google Gemini. 

Need help? Visit our Help Center to learn more.