Connected workspaces are high up on the wishlist for the majority of ChatGPT business users and OpenAI says it is listening.
OpenAI is already pushing hard to embed itself in Government departments, but this latest announcement shows it is also laser focused on becoming the go-to AI chatbot for all enterprise users.
The move will see “internal knowledge” rolled out for ChatGPT Team customers, focusing first on pulling in and using content from users’ Google Drive; which has itself had an AI glow-up in recent months.
ChatGPT Connecting to Internal Knowledge Sources
It was Nate Gonzalez, product leader at OpenAI who made the announcement on LinkedIn.
He shares that the OpenAI team hasbeen trying out the internal knowledge tool with Google Drive for several weeks. He says it has “…already changed how we onboard new team members, prep for meetings, and unblock work.”
This just in! View
the top business tech deals for 2025 👨💻
In a short demo video, the OpenAI team is shown using ChatGPT to find slides in Google Drive and “catch the user up” on a specific codenamed project using the new “internal knowledge” tab.
Beta Rolled Out but More To Come
The post also promises that Teams connectivity is coming “over the next few weeks”.
However, Gonzalez also asked for feedback from users who have been only too happy to share what connectivity they would like.
He teased: “…this is just the beginning. The team is already working on the next wave of connectors, aiming to support all the key internal knowledge sources your team relies on today—from collaboration and project management tools to data analytics platforms, CRMs, and more.”
But How Does This Work With Privacy?
Gonzalez says that he is most excited by the fact that the model will “learn your org’s unique language – project names and acronyms, and team-specific terms” but emphasizes that it will also “respect your user permissions”.
It’s of no surprise – though -that users have already raised concerns about data usage – after all this is a hot topic for all AI users. One commenter in LinkedIn asked: “What if ChatGPT access sources to add in own data and start giving same responses to others who are not in the same org?”
Gonzalez was quick to reply, stating: “We don’t train our models on any Team or Enterprise account data!” However – as Dropbox customers proved – there might still be concerns so this is probably a question that will we see being levelled again as more “connectors” are announced.