A globally coordinated system with the ability to immediately disable or restrict AI systems posing an imminent threat to human safety or societal stability. It's governed by an international body and triggered only under strict, pre-defined emergency conditions with multiple safeguards.