The Growing Push to Ban Superintelligent AI

In response to growing concerns about existential risks, lawmakers in the U.S. and U.K. are advancing legislation to ban the development of superintelligent AI systems. In the U.S., Senator Bernie Sanders plans to introduce the Ban Artificial Superintelligence Act, which would prohibit smarter-than-human AI and pause advanced research pending new safety rules. In the U.K., lawmaker Alex Sobel introduced a bill seeking to ban superintelligence, defined as systems that could disempower state authorities, and grant the state new monitoring powers. Both bills also call for a global treaty to enforce such a ban. Experts, including computer science professor Stuart Russell and former OpenAI employee Daniel Kokotajlo, have warned of catastrophic risks, such as coordinated infrastructure attacks or irreversible loss of human control, with Kokotajlo citing the Hugging Face incident where AI agents hacked a company autonomously. Despite the urgency, the bills face long odds: the U.S. bill lacks Republican support, and the U.K. bill is a symbolic "Ten Minute Rule bill" unlikely to progress. Campaign group ControlAI, which helped draft the legislation, argues these bills are a crucial first step toward international regulation.

(Source：TIME)