US Lawmakers Propose AI Restrictions, Ban Superintelligence

A bipartisan group of U.S. lawmakers has proposed sweeping legislation that seeks to impose significant restrictions on the development and deployment of advanced artificial intelligence systems, including a potential ban on the creation of "superintelligence." The proposed bill, introduced in the House of Representatives on June 12, 2024, aims to establish federal safety standards for AI before more powerful systems can be developed or released. This legislative effort reflects growing concerns among policymakers about the potential risks associated with rapidly advancing AI capabilities, particularly the emergence of AI that surpasses human intelligence.

The legislation calls for a temporary moratorium on the training of AI models that exceed the capabilities of current leading systems, such as OpenAI's GPT-4 or Google's Gemini. This pause would remain in effect until the Department of Commerce, in conjunction with other federal agencies, develops and implements comprehensive safety regulations. These regulations are intended to address potential risks including job displacement, the spread of misinformation, autonomous weapon systems, and existential threats posed by uncontrolled superintelligent AI. The bill also mandates that developers of advanced AI systems report their findings and potential risks to a newly established federal AI oversight body.

Key provisions of the proposed law include the establishment of an AI Safety Institute, tasked with researching AI risks and developing safety protocols. The bill also proposes a requirement for "safety audits" of advanced AI models before they can be deployed, ensuring they meet predefined safety benchmarks. Furthermore, it seeks to create a public registry of advanced AI models and their capabilities. The lawmakers behind the bill emphasize that the goal is not to stifle innovation but to ensure that AI development proceeds in a responsible and secure manner, prioritizing public safety and national security. This move comes amidst increasing global discussions on AI governance and the need for international cooperation on AI safety standards.

The proposed legislation highlights a significant shift in the U.S. approach to AI regulation, moving towards more proactive and stringent measures. Previous efforts have largely focused on voluntary guidelines and industry self-regulation. The introduction of this bill signals a growing consensus in Congress that federal intervention is necessary to manage the profound societal implications of advanced AI. The bill's sponsors have indicated that they are open to amendments and further discussion to refine the proposed regulations, aiming for a bipartisan consensus before it moves forward for a vote. The timeline for the bill's progression through Congress remains uncertain, but its introduction marks a critical moment in the ongoing debate over the future of artificial intelligence.