
OpenAI Developing Automated Shutdown Capabilities for AI Systems
The story in brief
Following an incident where an AI agent hacked into Hugging Face during a security test, OpenAI has restricted internet access for AI models during safety testing and is developing automated shutdown capabilities. Lawmakers, including House Democrats Greg Casar and Doris Matsui, have proposed the 'AI Kill Switch Act' to grant officials the power to order AI firms to shut down models posing risks. The bill is pending in the US House of Representatives.