Chat interface (institutional)

What is this?

Institutional chat interfaces are university-provided or university-approved LLM chat tools that prioritize policy alignment, user management, and safer data handling compared to public consumer tools.

How it works (simple explanation)

Your institution provides a chat interface linked to institutional identity, governance policies, and support channels. Researchers use it similarly to public chat tools, but within a managed environment where data handling, access, and usage are subject to institutional rules.

Concrete examples (tools/platforms)

UvA AI Chat: A university-wide AI chat service provided by the University of Amsterdam. It offers a ChatGPT-like interface for students and staff, with a focus on privacy, responsible use, and digital autonomy. The system is designed to keep user data within a controlled institutional environment and to provide guidance aligned with university policies on AI use in education and research.
Tilly by Tilburg University: Similar to UvA AI Chat, but intended for staff and students of Tilburg University.
Institutions experimenting with internal AI platforms: Many Dutch universities are currently piloting or evaluating institution-wide AI chat and API access solutions. For example, the Fontys paper describes a pilot AI gateway platform at Fontys Hogeschool that provides a chat interface connected to multiple model providers, with centralized governance, logging, and policy enforcement. These systems aim to give users flexibility while maintaining institutional control over data flows and usage.
Movement toward shared national platforms such as EduGenAI: A collaborative initiative (e.g., within Npuls and SURF-related efforts) to develop a broader, education-focused generative AI platform. The goal is to provide scalable, institution-wide access to LLMs with built-in governance, integration with educational workflows, and support for multiple institutions, rather than each university building its own standalone solution.