AI Safety Initiative

We don’t want anyone gambling with humanity’s existence.

We are not against today’s useful AI systems. We stand against the race toward systems that would escape human control. We want to prevent the creation of uncontrollable artificial superintelligence — and we have a plan for how to do it.

Read our plan Get in touch

Scroll down

Why now

A race no one is steering

Advanced AI is rapidly moving toward more autonomous and more general capabilities. If artificial superintelligence emerges before it is safe and under control, it may pose an existential risk to humanity. The field’s founders, Nobel laureates, researchers — and the very labs developing AI — have all warned about it.

The major AI companies publicly admit they would rather not create superintelligence — yet they are locked in a race that none of them can quit unilaterally. Only a common, enforceable agreement can help.

More about the risk and our goal →

Our goal

An international treaty by 2030

We are working toward a global, fully enforceable treaty that prevents the creation of uncontrollable and dangerous artificial superintelligence — with commitments from the key players, including the USA and China. Faced with global threats, humanity has reached agreements before. We want to build on them.

1970Treaty on the Non-Proliferation of Nuclear Weapons

1987Montreal Protocol on Substances that Deplete the Ozone Layer

2015Paris Climate Agreement

2030Treaty on artificial superintelligence — our goal

What we do

We are preparing a concrete plan

We are meeting with key stakeholders, shaping a realistic approach, and seeking funding to put it into practice.

Professionals, each in their own field

We are a team of academics, campaign strategists and technologists with experience in advocacy, the non-profit sector and the development of AI systems.

Ondřej Krása

Academia · AI risk

Head of the Department of Philosophy and Religious Studies at the University of Pardubice. Works on AI risk both academically and as a science communicator.

Amálie Kovářová

Campaigns · communications

Author of viral campaigns for Milion chvilek and Memory of Nations; long-time science communicator for the Czech Academy of Sciences.

Tamara Kováčová

Advocacy · lobbying

Lobbyist with more than a decade of experience; her advocacy work shaped Czech education and media law.

Martin Vraný

Technology · AI

Data scientist and developer of agentic AI systems; PhD in philosophy and cognitive science, research stay at UC Berkeley.

Full bios and why us →

For donors and partners

We are looking for founding partners

A relatively small founding gift today can decide whether Central Europe gets a professional organisation that moves governments, the EU and the public toward an international agreement on AI safety.

Become a partner