AI experiments and playgrounds

Why do we have AI experiments and playgrounds?

AI services change rapidly and new tools appear every month. Researchers, teachers, students and service units need ways to try these tools in real work, without turning every trial into a permanent service or creating unmanaged “shadow AI”.

AI experiments and AI playground solutions give the Aalto community controlled, time‑limited opportunities to:

• explore new AI tools and workflows,
• learn what works (and what doesn’t) in research, teaching and services,
• gather experience and feedback for future, more permanent solutions,
• reduce risks related to data protection, information security and costs.

They are designed so you can move quickly, but still stay within Aalto’s rules and support structures.

Three ways to work with AI at Aalto

At Aalto, you can work with AI in three main ways. The right choice depends on how mature your need is and how much stability you require.

1. Compliant AI tools

AI tools that Aalto already offers and supports for everyday work, teaching and research. These have been evaluated for data protection, information security, contracts and costs. You can usually start using them directly by following the guidance provided (for example, Aalto AI Assistant for general AI use).Approved compliant tools

2. AI playground solutions 

Physical or virtual environments where you can try AI tools hands‑on in a managed setup. For example, the AI Playground in ARTS offers local, AI‑ready high‑performance workstations for image generation, coding, prototyping, model testing and creative production. Playgrounds are shared spaces with preinstalled tools where you can explore without needing to acquire a new external service yourselfAI Playground | Aalto University

3. AI experiments

Temporary, small, controlled pilots with a new AI tool that Aalto does not yet provide as a standard service. An AI experiment has:

• a clearly defined purpose,
• a limited group of users,
• a limited period of use.

You propose the experiment, get Head of Department (or Service Director) approval and go through a light review. The goal is to learn quickly, not to create a permanent solution. Log in to see the details:

AI experimentation process in 2026 | Aalto University

When should you use which?

Use this guidance to decide where to start.

By choosing the right option, you help Aalto support safe, responsible AI experimentation while still enabling quick learning in a fast‑changing world.