AI WORLD HQ
You sit down to ask an AI assistant to draft a quick email. Twenty minutes later you're still pasting in background notes, summarizing old conversations, and trying to remember what the project was even about. According to a new report from Dropbox, that setup work — the part before the AI actually helps — is eating up more than a third of everyone's AI time.
What the report found
Dropbox's 2026 Trends Report, a piece of market research (a study the company runs to understand how people work) focused on AI's impact on content at work, found that the average person now spends 15.8 hours per week using AI tools. That's roughly two full workdays.
Of those 15.8 hours, 36.6% — or about 5.8 hours a week — goes to "gathering context." The full report includes more than 70 statistics about how AI is changing the way people handle content at work.
What "gathering context" actually means
In plain terms, context is everything the AI needs to know before it can do a good job. It's the previous emails, the project brief, the customer's history, the style guide, the meeting notes from two weeks ago.
When you ask an AI to "write a follow-up email to Sarah about the Q3 report," the AI doesn't automatically know who Sarah is, what Q3 report you mean, or what tone your team usually uses. Someone has to feed that in — and that someone is usually you.
So the 5.8 hours a week is the time workers spend digging through files, copying and pasting, summarizing, and re-explaining things to the AI before the actual task can begin. It's the overhead that nobody counts when they brag about how much time AI saves them.
Why this matters
The numbers suggest AI isn't quite the time-saver it's often sold as — at least not yet. If you're spending 15.8 hours a week with AI and 5.8 of those hours are prep work, the "real" output time is closer to 10 hours a week.
That's not a reason to stop using AI. But it is a reason to be honest about where the time goes — and to look for ways to cut down on the setup. Clearer prompts (the instruction you give to the AI), saved templates, and tools that remember your past conversations can all shrink the context-gathering part.
Wrap-up
Dropbox's report puts a number on something most AI users feel but rarely measure: a lot of "AI time" is actually setup time. The good news is that the setup is learnable. Pick one task, write down the context it needs, and try it again next week. You'll probably find that your 15.8 hours start producing more output — and the AI starts feeling less like a tool you wrestle with and more like one that actually helps.