Access Permissions and Data Privacy

Giving an agent access to your tools is what makes it genuinely useful – but it also means you are sharing data with an external system. Understanding what you are sharing, with whom, and how to control it is not a technical skill. It is a basic professional responsibility, the same way knowing who has access to your shared drives is.

What Happens to the Data You Share

When you paste text into an agent or upload a document, that content is sent to the platform's servers for processing. What happens to it after that depends on the platform and the type of account you are using.

Most consumer accounts – free or personal paid plans – allow the platform to use your conversations to improve their models unless you explicitly opt out. Most business and enterprise accounts include contractual data privacy protections that prevent this.

The Three Questions to Ask Before Sharing Data

Before you paste a document or connect a tool, ask yourself:

• Is this information confidential? If it contains personal data, financial details, client information or anything marked internal or confidential, check your organization's AI policy before sharing it;
• What account am I using? A personal free account has different privacy terms than a company enterprise account. Know which one you are on;
• Does this platform need this access? Only connect tools the agent will genuinely use. Connecting your entire Google Drive when you only need it to read one folder is unnecessary exposure.

Managing Permissions Practically

Most platforms make it straightforward to review and revoke access. It is worth spending five minutes doing this audit when you first set up your agent, and then revisiting it every few months.

What to check:

• Which tools and accounts are connected to your agent;
• Whether your conversation history is being used for model training and how to opt out if you prefer;
• Whether you are using a personal or a business account, and what the data terms are for each.

What is the difference between consumer and enterprise AI accounts?

Consumer accounts – including free plans and most individual paid subscriptions – are designed for personal use. They typically include the platform's standard data terms, which may allow conversation data to be used for model improvement unless you opt out.

Enterprise accounts are designed for organizational use and typically include:

• a data processing agreement (DPA) with the provider;
• zero or limited data retention after sessions close;
• no use of your data for model training;
• admin controls that let your IT team manage access across the organization.

If you regularly work with sensitive client or company data, using an enterprise account – or your organization's approved AI tool – is the appropriate choice, not a preference.