Munich's community for making AI safe

We are a group of students and professionals, learning about AI safety and working on it together.

Where to start

We run two student groups at TUM and host events anyone can join. New to the field? Start with our reading list.

Start here

Fundamentals

A self-guided introduction to the core questions: why advanced AI could cause serious harm, what researchers are doing about it, and where you might fit in. Read and watch at your own pace, then bring your questions to an event.

Open the reading list
TUM Students

Compute Verification

For any international agreement on AI to work, countries need to be able to check that others are keeping to it. The technology to do this isn't ready yet, and we're growing the community that can help build it.

Branch details
TUM Students

AI Governance

We aim to create transparency around the regulatory guidelines for AI development, and around the gaps we see in how AI companies report on safety.

Branch details

Why AI safety?

Without adequate safeguards, powerful AI could autonomously escape our control, help bad actors build weapons, or slowly take important decisions out of human hands. We research these problems and help students start working on them.

100+

experts contributed to the 2026 International AI Safety Report, chaired by Turing Award winner Yoshua Bengio, with advisers nominated by more than 30 countries and international organizations.

Source: International AI Safety Report 2026 (opens in a new tab) (February 2026)

5%

is the median probability that nearly 2,800 AI researchers gave to advanced AI causing human extinction or a similarly severe, permanent loss of human control.

Source: AI Impacts, 2023 Expert Survey on Progress in AI (opens in a new tab) (published January 2024)

<10 years

until superhuman AI, according to the top AI forecasters in the world.

Source: AI Futures Project (opens in a new tab) (Aug 2026)

Come to an event

Most events are open to everyone, some are for students only.