Anthropic proposes a bill allowing Washington to ban dangerous AI models.

Anthropic Proposes Legislation to Regulate AI Models

Anthropic, the developer of the AI assistant “Claude,” confirmed that it is negotiating a bill with U.S. lawmakers aimed at regulating the artificial intelligence sector. This move comes in response to the absence of comprehensive federal legislation addressing the risks associated with rapidly advancing technology.

Key Provisions of the Proposed Bill

According to the draft legislation, the U.S. government could prohibit the launch or use of any AI model if it poses an imminent risk that could lead to the death or serious injury of more than 50 individuals or cause damages exceeding one billion dollars. This action would require judicial approval.

The company stated that the proposal grants the federal government the authority to halt or deter dangerous deployments of powerful AI models. Additionally, it mandates that developing companies conduct advanced safety tests, transparently disclose their results, submit models for independent evaluations, and maintain robust security programs.

Current State of AI Regulation in the U.S.

Currently, AI regulation in the United States relies on voluntary commitments from companies and executive orders from the government. To date, Congress has not passed any comprehensive federal law to regulate the sector.

Calls for stricter oversight of AI technologies have intensified following incidents involving “AI agents,” systems capable of autonomously executing sequential tasks. These events have raised growing concerns about potential security risks.

Federal Authority and State Legislation

The proposed legislation includes provisions that grant the federal government extensive powers that exceed some existing state laws. This step aims to unify the regulatory frameworks for artificial intelligence across the United States.

Supporters of the bill argue that it provides a legal framework to address the potential risks of advanced models. However, some experts view it as less stringent than other proposed bills aimed at regulating advanced AI technologies.

To follow the news in Arabic