If AI-generated writing has ever felt a little too polished—or simply not like you—ChatGPT Work is getting a feature that could change that.
OpenAI is rolling out a personalization capability that allows ChatGPT Work to learn how you write by using information from connected apps and files. Instead of relying entirely on generic instructions such as “make this sound casual,” ChatGPT can pick up the patterns that make your writing recognizable.
That includes things like phrases you regularly use, how you sign off messages, sentence structure, and even capitalization habits.
ChatGPT Can Learn Your Personal Writing Style
ChatGPT Work can now pick up on what makes your writing sound like… you. Your favorite phrases. Your very specific sign-off. your capitalizations quirks.
Connect the tools you use every day, like Gmail, Google Drive, Slack, and SharePoint, and ChatGPT Work will learn your… pic.twitter.com/jiP9K4xhy5
— ChatGPT (@ChatGPT) September 7, 2026
The idea is fairly straightforward: give ChatGPT access to the tools you already use for work, and it can use that context to produce writing that better matches your established voice.
ChatGPT Work is designed to work across connected apps and files, rather than functioning only as a traditional chatbot. OpenAI describes Work as an agent capable of researching information, working across apps and files, and producing finished documents and other work products.
With this personalization capability, the AI can potentially recognize patterns such as:
- Words and phrases you frequently use
- Your preferred level of formality
- How you start and end messages
- Sentence length and structure
- Capitalization and formatting habits
- The overall tone of your emails and documents
The result should be drafts that require less rewriting before you can actually send or publish them.
Gmail, Slack, Google Drive and SharePoint
The feature becomes particularly interesting because ChatGPT Work can operate across connected workplace services.
OpenAI supports apps that connect ChatGPT with services such as Google Drive and Slack, allowing ChatGPT to search and reference information from those platforms.
Google Drive, for example, can provide access to Google Docs, Sheets and Slides that the connected account is already authorized to access.
Microsoft environments can also connect through the SharePoint app, which lets ChatGPT search and reference SharePoint and supported OneDrive work or school content subject to permissions and workspace settings.
That gives ChatGPT a much larger pool of context than a single conversation.
Why This Could Make AI Writing Better
One of the biggest problems with AI writing isn’t grammar or spelling. It’s voice.
A generated email can be perfectly written and still sound obviously AI-generated because it doesn’t use the same expressions or communication style as the person sending it.
Personalization addresses that problem by allowing ChatGPT to learn from real examples.
For someone who regularly writes customer emails, project updates, reports or internal messages, the difference could be significant. Instead of repeatedly explaining your preferred tone, you could let ChatGPT learn those preferences from your existing work.
OpenAI’s Work experience already supports creating and editing documents, spreadsheets, presentations and reports from instructions and source material, including matching an existing file or template.
Adding a learned writing voice makes that workflow considerably more personal.
But Privacy Is the Big Question
There is an obvious trade-off.
The more context ChatGPT has about how you communicate, the better it can personalize its responses—but that also means users and organizations need to understand what information ChatGPT can access and how those connections are controlled.
OpenAI says connected apps only provide access according to the permissions granted to the connected account and workspace settings. Users can also disconnect connected Google apps from ChatGPT settings.
For organizations, administrators can control app availability and permissions. OpenAI also says app calls in managed workspaces are logged through its Compliance Logs platform.
That’s important because workplace writing can contain sensitive customer, financial, legal or internal information.
How to Get Started
For eligible users, connected apps can be configured through ChatGPT’s settings. OpenAI’s current app setup flow allows users to connect supported services and complete their authentication and permission steps.
ChatGPT Work itself is available on web and mobile for eligible paid plans, with availability depending on the plan and workspace configuration.
Once the relevant apps are connected, the personalization experience can use authorized information to make future work more consistent with your style.
ChatGPT Is Becoming More Like a Personal Work Assistant
This feature points toward a broader shift in how OpenAI is positioning ChatGPT.
The goal is no longer simply to generate text from a prompt. ChatGPT Work is increasingly designed to understand the context surrounding the work, access relevant information, and produce something closer to a finished result.
Learning a user’s writing voice is a natural extension of that approach.
The potential payoff is simple: less time fixing AI-generated text because it already sounds like something you would write.
The bigger question will be whether users are comfortable giving an AI access to enough of their digital workspace to make that personalization genuinely useful.
For continuing updates on OpenAI, ChatGPT, Codex, GPT models, and AI industry developments, bookmark:
