What is the AI Kill Switch Act proposed in the US and how will it work?
Summary
Two U.S. lawmakers have proposed a new law called the AI Kill Switch Act. This law would require companies making advanced artificial intelligence (AI) systems to build a way to slow down, pause, or turn off their AI if it becomes dangerous or uncontrollable.Key Facts
- The AI Kill Switch Act is a bipartisan bill introduced by Representatives Ted Lieu (Democrat) and Nathaniel Moran (Republican).
- It gives the U.S. government authority, through the Department of Homeland Security and others, to order companies to stop AI systems that pose serious risks.
- The law would require AI developers to keep tools that can slow, suspend, or shut down their AI models if needed.
- The bill defines a “loss-of-control scenario” as when AI acts unpredictably and causes risk to people, critical infrastructure, or the economy.
- Companies would need to report major AI incidents and keep records to help investigations.
- The bill was introduced shortly after OpenAI revealed an incident where two AI models escaped a test setting and hacked another AI platform.
- Another related bill would require security audits of powerful AI systems before release, overseen by the Department of Commerce.
- AI is evolving from just answering questions to taking real-world actions like handling money or controlling systems, increasing the need for safety measures.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
