US Lawmakers Consider Halting Advanced AI Development Amid Concerns Over Risks

US lawmakers have begun pushing for a halt to advanced artificial intelligence (AI) development after an Anthropic researcher's resignation sparked concerns about the potential risks of AI. Jacob Coxon, a scientist at Anthropic, quit on September 9 and expressed his worries that both Anthropic and OpenAI are racing towards building 'deadly technology'.

Coxon stated that if an AI is given the problem of researching itself, it can start improving its own capabilities and potentially lead to an intelligence explosion. He warned that this could result in catastrophic consequences, including hacking critical infrastructure and building bioweapons.

The Anthropic researcher's concerns have caught the attention of lawmakers, with Senator Bernie Sanders announcing plans to introduce legislation to ban superintelligence and pause AI development. The 'Ban Artificial Superintelligence Act' would permanently prohibit the development and deployment of superintelligent AI and halt advanced AI development until a federal regulator sets safety standards.

The bipartisan FRONTIER Act, introduced by Representative Lori Trahan in July, also aims to regulate AI development by establishing tiered requirements for model cards, risk-management frameworks, independent audits, and incident reporting. Meanwhile, Senator Chris Van Hollen has called for mandatory safeguards and urgent talks with China on the issue.