Thesis project: Shadow AI
Catch shadow AI on the wire. Law firms handle highly confidential case files, contracts and correspondence. At the same time, employees increasingly use generative AI through personal accounts to draft or summarise faster. That usage falls outside standard data-loss-prevention controls, and smaller firms have no security team watching for it. You test whether a lightweight, network-level detection approach can restore that visibility.
- Thesis project
- MSc
- 1 place
- Amsterdam
About Datagrove
Datagrove is an AI-native development agency on the Herengracht in Amsterdam. We build AI systems, platforms and apps for businesses, and we regularly speak with law firms that want to use AI without confidential files leaving the office.
Enterprise AI-governance platforms are unaffordable for those firms. This research tests whether a lightweight, network-based approach can close the same blind spot.
What you do
The research question: to what extent can network analysis reliably detect employees at a law firm pasting confidential client data into unmanaged, consumer-grade AI tools, in a representative small-firm network environment? The proposed method:
- Set up a controlled lab environment simulating a small law firm's network.
- Identify the traffic and content patterns associated with pasting text into consumer LLM interfaces.
- Test detection accuracy against a set of simulated scenarios.
- Evaluate the approach on accuracy, false-positive rate and practical deployability for a firm with no dedicated security staff.
Who we are looking for
This project is listed on the UvA Thesis Matching Platform and is open to master's students of these programmes:
- MSc Computational Science (joint degree).
- MSc Information Studies, Data Science track.
- MSc Security and Network Engineering.
- You understand network traffic and want to build a lab and measure, not just read literature.
- Start as soon as possible. 1 January 2027 works too, when most programmes begin their thesis semester.
What we offer
You research a problem law firms bring to us themselves, with a test environment built around it. Research is not a side activity for us: we want to keep learning ourselves, and this project is part of that.
- A lab-like test network environment, informed by conversations with prospective law-firm clients about their data-safety requirements.
- Daily supervision by Sion Coolwijk, UvA alumnus MSc Data Science, with hands-on experience deploying AI systems for businesses.
- Freedom in how you carry out the research, guided by a daily supervisor. The research question is a starting point, not a fixed frame: if another domain or method proves stronger along the way, there is room to adjust the direction.
- A desk on the Herengracht in Amsterdam.
Apply for this role
Leave your details, write a short motivation and attach your resume. We read every application ourselves and respond quickly.