Skip to main content

ChatGPT, Copilot or your own AI?

Three routes compared, plus the business accounts for ChatGPT, Copilot, Gemini and Claude, a decision guide and a checklist for approval.

9 min read

If you use Microsoft 365, the question comes up quickly: is Copilot enough, do you also need ChatGPT, or your own application straight away? The answer depends less on the tool than on three features of the task: how much internal knowledge it needs, how often it comes up and where the documents are kept today.

Whether a route is compliant with data protection law then depends mainly on the contract, the settings and the rules. That applies to all three routes.

General AI assistants

ChatGPT, Claude and Gemini are versatile: writing drafts, summarising, translating, collecting ideas. They know your company only from what you type into the prompt or upload. To use them in a team, you need a business account with a suitable contract.

For ChatGPT Business, ChatGPT Enterprise and the API platform, OpenAI states that it does not use business data for training by default. For these three, OpenAI offers a Data Processing Addendum, in other words a data processing agreement (DPA) under Article 28 GDPR. According to OpenAI, data from its versions for individuals is also used for training.

Two points need settling before rollout. In ChatGPT Business, workspace admins can view, export and delete users' conversations. Tell your team openly who can do this and for what purpose. Admins also decide how long conversations are retained. According to OpenAI, deleted conversations are removed within 30 days, apart from legal and a few other exceptions.

Copilot in Microsoft 365

If your company works with Microsoft 365, Copilot can access documents, emails and calendar entries with the paid Copilot licence, using the permissions of each individual user. For this, permissions must be set up properly: whatever someone is allowed to see, Copilot will find too. Microsoft states that Copilot only surfaces organisational data to which the individual user has at least view permissions.

On data protection, Microsoft states the following: prompts, responses and data accessed through Microsoft Graph are not used to train the underlying language models. For users in the EU, traffic stays within the EU Data Boundary. Models from Anthropic, which are provided as sub-processors, are currently excluded from this. Administrators decide whether such third-party models are used. Microsoft has since renamed the product “Microsoft Copilot”, and some licences still show the old name.

A simple test before launch: a person without special permissions asks Copilot about salaries, personnel files or dismissals. Whatever comes up is already findable for that person today. Restrict such sharing in SharePoint, OneDrive and Teams before Copilot is rolled out to everyone.

Business accounts compared: ChatGPT, Copilot, Gemini and Claude

ChatGPT, Copilot, Gemini and Claude each come in several account types. For work in a company, only the business ones are an option: ChatGPT Business (called ChatGPT Team until August 2025) or Enterprise, Copilot in Microsoft 365, Gemini in Google Workspace, and Claude Team or Enterprise. They differ less in the model than in who manages them, which data they can access and what the contract says. We leave out prices because they change often and are listed on the providers’ websites.

OpenAI

ChatGPT Business

Formerly ChatGPT Team. For small and growing businesses, from 2 seats

A shared workspace for your team. Admins set how long conversations are kept and can view, export and delete the conversations in the workspace.

Training
not on your data by default
Contract
data processing addendum via an OpenAI form

help.openai.com

OpenAI

ChatGPT Enterprise

For companies rolling out ChatGPT across the whole organisation

On top of that, controls for the whole organisation and an audit log of conversations via the Compliance API. Admins decide how long data is kept.

Training
not on your data by default
Contract
data processing addendum via an OpenAI form

openai.com

Microsoft

Microsoft Copilot Chat

Included in eligible Microsoft 365 licences, no add-on licence needed

Works mainly with the web and with what you upload or paste. In Outlook, Copilot Chat also draws on emails, calendar, meetings and chats, to a limited extent.

Protection
enterprise data protection
Your data
only in Outlook and what you share yourself

learn.microsoft.com

Microsoft

Microsoft 365 Copilot

Add-on licence per person, for anyone who searches a lot in email, Teams and documents

Uses Microsoft Graph to access the emails, meetings, calendars, chats, documents, sites and files the person has rights to, and works directly in Word, Excel, PowerPoint and OneNote.

Your data
everything the person is allowed to see
Beforehand
check sharing in SharePoint, OneDrive and Teams

learn.microsoft.com

Google

Gemini in Google Workspace

Included in Google Workspace plans

Gemini works in the side panel of Gmail, Docs, Sheets, Slides, Drive and Chat with your emails and documents, takes notes in Meet and is available to everyone in the account as the Gemini app.

Training
not without your permission or instruction
Contract
prompts are customer data under the Cloud Data Processing Addendum

knowledge.workspace.google.com

Google

Gemini Enterprise

A separate Google Cloud product, Business edition for 1 to 500 users

Intranet search, AI assistant and platform for AI agents. Connects to sources such as SharePoint, Confluence, Jira and ServiceNow, and you can run your own agents on it.

Access
answers only from content the person is allowed to see
Getting started
Business edition in the browser

docs.cloud.google.com

Anthropic

Claude Team

For teams of 2 to 150 people

Central administration and billing, sign-in via SSO and admin controls for connectors. Claude Code is included.

Training
not on your content by default
Contract
data processing addendum is part of the commercial terms

claude.com

Anthropic

Claude Enterprise

From 20 seats

Adds automatic user provisioning (SCIM), roles with fine-grained permissions, audit logs, the Compliance API and custom retention periods.

Training
not on your content by default
Contract
data processing addendum is part of the commercial terms

claude.com

How to find the right account:

  • If your team works with Microsoft 365, Copilot Chat is already included in many licences. Start there and give the add-on licence first to the people who search a lot in email, Teams and documents.
  • If your team works with Google Workspace, Gemini is already included in the plans. First check which Gemini features are enabled in your admin console.
  • If you use neither Microsoft 365 nor Google Workspace, or want a second assistant, ChatGPT Business or Claude Team is enough for smaller teams.
  • If you need an audit log, automatic user provisioning or custom retention periods, for example because of professional secrecy or internal audit, the Enterprise accounts are the ones to consider.

The checklist further down applies to every account type. Information as of October 2026.

Your own AI application

Your own application is tailored to one task, such as drafting quotes from your templates or answering new colleagues' questions from the staff handbook. It works with your knowledge, can show the source of its answers and can be run in the EU if you wish.

Such an application often combines a language model with a collection of your own documents. The technical term is retrieval-augmented generation, or RAG. The German Data Protection Conference (DSK) published an orientation guide on this in October 2025. According to the guide, established access rights and role concepts can be applied in the document collection, but not in the language model itself. The method can reduce incorrect answers to a certain extent and in many cases makes it possible to run the language model in-house.

If you have such an application developed and use it under your own name, you count as a provider under the EU AI Act. For most applications this means few additional obligations. It still belongs in the planning.

If the application should not just draft texts but carry out steps in your programs, it is called an AI agent. How much it may do on its own is covered in the article AI agents in business.

How to decide

General assistantwhen many different tasks come up and little internal knowledge is needed
Copilotwhen you use Microsoft 365, permissions are set up properly and it is clear which models are approved
Your own applicationwhen a task comes up often, needs your knowledge and has to run reliably

How to assign a specific task:

  • If the task needs no internal documents, such as a draft for a specialist article, a general assistant is enough.
  • If the documents you need are already in Outlook, Teams or SharePoint and the permissions are right, Copilot saves you uploading them.
  • If the same task comes up every week, uses fixed templates and the result has to be traceable, consider your own application.
Two routes often run side by side: an assistant for everyone and your own application for the task that takes the most time.

Example: one quote, three ways

A planning consultancy with 15 employees writes several quotes every week. They are based on an enquiry by email, standard text blocks for services and previous quotes.

With a general assistant, someone copies the enquiry without the customer's name, together with the relevant text blocks, into the prompt and asks for a draft. That works. But the quality depends each time on which blocks the person picks and includes.

With Copilot, the prompt might read: “Draft a quote based on today's enquiry and our ‘Planning quote’ template.” Copilot finds the email and the template itself, provided the person has access to them. The quality depends on how tidy the file storage is.

With your own application, the team selects the enquiry. The application knows all approved text blocks, suggests the right ones and shows which template each paragraph comes from. The development effort can pay off because the task comes up every week.

In all three cases a person checks the draft before it leaves the company.

Checklist before approval

These questions clarify before approval whether a tool suits your organisation, whether it is an assistant, Copilot or a development partner:

  • Is there a business account that we can manage centrally and block when employees leave?
  • Is there a data processing agreement, and does it cover exactly the plan we want to use?
  • Who in our organisation can view, export or delete conversations, and does the team know?
  • Which models are approved, including those from third-party providers?
  • For Copilot, also: have the sharing settings in SharePoint, OneDrive and Teams been checked?

The questions on training, data location, retention and sub-processors are in the checklist in the article Data processing agreements with AI providers. Record the answers for each tool with a date. When the provider makes changes, you can then see what needs checking again.

Next steps

  • Write down the five tasks your team would use AI for most often and assign each one to an option using the three “if” questions.
  • Work through the checklist for the option you have chosen.
  • If Copilot is an option, run the permissions test with a person who has no special permissions.
  • If a task looks like a case for your own application, describe it on one page: input, documents, desired result, who checks it.

Which data belongs in a prompt is covered in the article AI and the GDPR: which data you may enter.

Worksheet as PDF: Checklist for AI tools

The templates and checklists from this article to print and fill in. You get the PDF when you subscribe to the lindenbrook newsletter with new articles and worksheets. The newsletter is in German.

By clicking, you subscribe to the newsletter. You will receive an email straight away. The link in it confirms your subscription and takes you to the PDF. You can unsubscribe via the link in every issue. The newsletter is sent via bemailed, which measures opens and clicks. More in our privacy policy.

Frequently asked questions

Do we need all three?

No. Many start with an assistant for everyone and add their own application for the one task that takes the most time.

How long does your own application take?

A first draft that your team can use itself is created in a two-day workshop. Afterwards you know whether the approach holds up.

Is an individual ChatGPT account enough for people in the team?

Not for working with personal data. According to OpenAI, it offers a data processing agreement for ChatGPT Business, ChatGPT Enterprise and the API. OpenAI also uses data from its versions for individuals for training.

Is ChatGPT Business the same as ChatGPT Team?

Yes. OpenAI renamed ChatGPT Team to ChatGPT Business on 29 August 2025. A Business workspace needs at least two seats.

What is the difference between Copilot Chat and Microsoft 365 Copilot?

Copilot Chat is included in eligible Microsoft 365 licences and works mainly with the web and with what you upload. The add-on licence Microsoft 365 Copilot uses Microsoft Graph to access the emails, meetings, chats and documents the person is allowed to see, and works directly in Word, Excel and PowerPoint.

Is ChatGPT Business enough, or do we need Enterprise?

For most small and medium-sized teams, ChatGPT Business is enough. Enterprise is worth it if you roll out ChatGPT across the whole organisation and need, for example, an audit log via the Compliance API. OpenAI offers the data processing addendum for both.

What do the business accounts cost?

Copilot Chat is included in eligible Microsoft 365 licences, and Gemini in Google Workspace plans. The providers charge for the other accounts per seat, and current prices are on their websites. We don’t give amounts here because they change often.

Is Gemini included in Google Workspace?

Yes. According to Google, the Gemini app and Gemini in Gmail, Docs, Meet and other apps are part of Google Workspace plans. Google does not use your prompts for training without your permission or instruction. Gemini Enterprise, by contrast, is a separate Google Cloud product for search and AI agents across many sources.

Are ChatGPT Business, Copilot and Claude Team GDPR compliant?

That doesn’t depend on the account alone. OpenAI and Anthropic offer a data processing addendum for their business accounts, Microsoft refers to its Product Terms and Data Protection Addendum for Microsoft 365 customers, and Google to the Cloud Data Processing Addendum. According to all four, they don’t train on your inputs by default. Whether your use complies with data protection law then comes down to your settings and rules: which data may go in, who can see conversations and how long they are kept.

Does Anthropic train on our data in Claude Team?

According to Anthropic, no: in Claude Team and Claude Enterprise, your content is not used for training by default. The data processing addendum is part of the commercial terms.

Sources

  1. Microsoft: Data, privacy and security for Microsoft 365 Copilot
  2. Microsoft: Copilot overview
  3. OpenAI: Enterprise privacy
  4. Orientation guide on generative AI systems using the RAG method, German Data Protection Conference (DSK), October 2025 (in German)
  5. OpenAI: ChatGPT Business overview
  6. Anthropic: Plans and pricing
  7. Anthropic: Data processing addendum (DPA)
  8. Google: Generative AI in Google Workspace Privacy Hub
  9. Google: Google Workspace with Gemini
  10. Google Cloud: Gemini Enterprise overview
  11. Google Cloud: Gemini Enterprise editions

Updated: . This article is not legal advice.

From idea to first applicationTwo days · one of your tasks becomes a draft you can operate yourself
See the workshop

Read next

Questions about your situation?

Write to us. We will get back to you personally with a proposal for the next steps.