Your data stays yours. In writing.

Last updated October 21, 2026.

This page explains, in plain English, how Modern Thesis handles client data. It describes how we work. The written agreement for each engagement sets the specific terms.

What data do you access, and why?

We access the systems and records that the workflows in scope run on, such as shared documents, spreadsheets and the software a process runs through. In the assessment, we use that access to map how work moves and measure a baseline. In the build, it lets us connect AI to your systems, and in stewardship, it lets us monitor accuracy and keep those systems running. We don't access data that the agreed work doesn't need.

Do you use our data to train AI models?

No. We never use client data to train AI models, ours or anyone else's. Your data is used only to build and run your own systems. It is not pooled with other clients' data, and it is not used to improve anything we sell to anyone else.

When is access agreed?

Before we start. We agree in writing how your data will be handled before we access anything: which systems, which records, who on our team will have access and for what purpose. If the work later needs access to something new, we agree that in writing too, before we use it.

How much access do you need?

Only what the work needs. We ask for the narrowest access that lets us do the job, such as one system instead of all of them, or read-only access where we don't need to change anything. You can see what access we have and remove it at any time.

Who owns the data, accounts, code and documentation?

You do. Your data stays yours. The accounts your systems run on, the code we write for you and the documentation that explains how it works all belong to your company. If you ever want to bring the work in-house or hand it to another provider, you have everything you need to do that.

Which AI providers do you use?

We choose AI providers whose business terms don't train on customer data. When a system we build sends information to an AI model, we check the provider's terms before we use it, and we tell you which providers your systems rely on. Where the tools you already pay for meet that standard and fit the work, we use them.

Where do people check what the AI does?

Wherever judgment matters. We design human review into the workflows we build, so a person checks AI output before it reaches a client, a decision or a record that counts. During stewardship, we measure accuracy over time and adjust the system when it slips.

What happens when an engagement ends?

We remove our access. When an engagement ends, our team's access to your systems and data is removed, and you keep the accounts, code and documentation. We can walk your team or your next provider through the handover.

What if our business has specific confidentiality rules?

Tell us in the first conversation. Professional services firms, for example, often have duties to their own clients about how information is handled. When we know your requirements before we start, the written agreement can address them.

How Modern Thesis handles this

We settle data handling in writing before any work begins, and we raise it in the first conversation. If you have questions about this policy, or want to walk your team through it, email Jim at jim@modernthesis.org.