Google Docs users who want the "Write with Gemini" prompts out of their face can switch off the feature by changing a pair of settings inside Google Workspace. That's the practical news. The larger point is harder to miss: the AI layer is now being threaded into routine office software whether people asked for it or not.
According to the reported steps, users need to open Google Docs, click the gear-shaped Settings icon at the top right, then go to Preferences and turn off Smart features and personalization in Google Workspace. They also need to switch off Smart features and personalization in other Google products. Once those are disabled, the Gemini writing prompts should disappear.
It's a small consumer how-to story, yes. But it says something bigger about the state of AI software in 2026. Tech companies keep presenting these assistants as obvious upgrades. A lot of users experience them as clutter.
Key Facts
- The reported change applies to Google Docs and its "Write with Gemini" prompts.
- The source report was published on June 17, 2026.
- Users must open the Settings menu in Google Docs, then go to Preferences.
- Two settings need to be turned off: smart features in Google Workspace and in other Google products.
- The source article came from TechCrunch's technology coverage.
Where Google put the off switch
The instructions are straightforward once you know them, which is not the same thing as being obvious. Inside Docs, the relevant controls sit under Preferences rather than under anything labeled Gemini or AI. That's a familiar pattern in modern software design: the feature arrives with a big button, while the refusal is buried a layer or two down.
And no, that isn't an accident. Friction matters. If enabling is easy and disabling takes a hunt, companies get better adoption numbers.
For readers who don't track this stuff closely, Gemini is Google's family of AI models and products, used across search, phones, and workplace tools. A large language model, in one clean sentence, is software trained on huge amounts of text so it can predict and generate plausible language. That can be useful. It can also be wrong, intrusive, or simply annoying when all you wanted was to draft a meeting note without a chatbot leaning over your shoulder.
The fight over AI in office software isn't about capability anymore; it's about consent.
Google has been methodically adding Gemini across its productivity suite, part of the same broad push that has turned almost every major software product into an AI showroom. We've seen the industry sell this as inevitability for months. But a product launch isn't a breakthrough, and a pop-up in a blank document doesn't become helpful just because the model behind it is expensive.
Why this lands differently in office software
Search has always been noisy. Social platforms are built on interruption. A word processor is different. People open Docs to think, draft, revise, and collaborate. That work depends on rhythm and concentration, which is exactly why unsolicited prompts feel more invasive there than in a feed-based app.
Still, Google's strategy is plain enough. If people won't always go looking for AI tools, put the tools directly in the path of ordinary work. Microsoft has done its version of this. OpenAI would like every screen to become an entry point. The pressure is industry-wide, and the sales pitch is familiar: you'll be faster, smarter, more productive. Sometimes you are. Sometimes you've just acquired one more blinking suggestion to ignore.
That tension runs through other policy fights too. In consumer tech, companies increasingly present defaults as destiny, then frame opt-outs as a niche concern. The same logic sits behind arguments over age-gating and platform design in Britain, where regulators keep discovering that defaults shape behavior far more than slogans do, as we saw in the UK's proposed under-16 social media restrictions and the loopholes examined in BreakWire's look at the major gaps in that plan.
Google, for its part, has publicly framed Workspace AI as an assistance layer rather than a replacement for human work. The company describes Workspace features on its own product pages, and broader company information on Gemini is available through Wikipedia's entry on Gemini and Google Workspace. But product language and user experience are not the same thing. Users tend to judge by interruption, not architecture.
The opt-out problem
Here's the thing: an off switch is only meaningful if people can find it, understand what it changes, and trust that it sticks. Tying Gemini prompt removal to broader smart-feature and personalization settings may solve the immediate annoyance, but it also raises a basic design question. Why should a user who wants fewer AI nudges have to toggle a wider personalization bundle to get there?
That's where the industry's "AI everywhere" phase starts to look less like empowerment and more like bundling. Companies aren't just offering a tool. They're wrapping it into settings that govern multiple behaviors, which makes the choice messier than it needs to be.
And this isn't some abstract consumer-rights seminar. In workplaces, software defaults shape what employees see every day. If IT departments leave the settings on, staff may assume the prompts are mandatory or baked in too deeply to change. If individuals turn them off, they may still wonder what else those settings affect. Confusion is a feature of bad product design, even when companies call it integration.
For readers trying to make sense of the broader AI push, that matters more than another keynote promise. We already know tech chiefs want students and workers to absorb AI tools into daily life; they keep saying so, as in BreakWire's coverage of AI advice from Google and Nvidia leaders. What we know less clearly is how many people actually want these systems threaded through every familiar interface by default.
There are also plain privacy and data-use questions attached to personalization settings in large platforms. Users looking for official background can read Google's support and product materials, while broader digital-governance standards are regularly discussed by bodies such as the U.S. National Institute of Standards and Technology's AI Risk Management Framework and international agencies including the UNESCO recommendation on AI ethics. Those frameworks won't tell you how to dismiss a Docs prompt. They do explain why defaults, transparency, and user control keep coming up.
What to watch from here
The immediate next step for users is simple: open Docs, head to Settings, enter Preferences, and switch off both smart-feature personalization controls if you want the Gemini prompts gone. The more interesting next step is Google's. If enough users start hunting for escape hatches, the company will have to decide whether to offer a cleaner, explicit Gemini toggle inside Docs itself.
Watch for any update to Google Workspace settings pages or support documentation after the June 17, 2026 report, because that will show whether Google treats this as a minor usability complaint or as a sign that the AI-first interface still hasn't won over ordinary document writers.