A group of four senators, led by United States Senate Majority Leader Chuck Schumer, is proposing a plan for Congress to allocate a minimum of $32 billion over the next three years to advance artificial intelligence (AI) and establish protective measures around it.
This roadmap is part of the U.S. government’s ongoing efforts to regulate and promote the development of AI. It comes just six days after U.S. legislators introduced a bipartisan bill to support President Joe Biden’s administration in implementing export controls on top AI models created within the country.
After extensive consultations with industry experts and critics of AI, the bipartisan working group recognized the significance of investing in AI to maintain America’s competitiveness with global rivals and enhance the quality of life for its citizens. They emphasized the potential of AI technology to contribute to medical breakthroughs, such as finding cures for cancer or chronic illnesses.
While the roadmap is not a concrete legislative proposal, it offers insights into the scale and importance of future AI regulations, laying the groundwork for more comprehensive and detailed policies to follow.
The senators’ proposal also emphasizes the need to enforce existing laws for AI, including addressing any gaps or unintended biases that may arise. They advocate for the development of testing standards to understand potential harms caused by AI and recommend specific requirements for transparency and explainability in different use cases.
Furthermore, the group suggests implementing new transparency requirements as AI products are introduced to the market and conducting studies to assess the impact of AI on jobs and the American workforce.
In parallel to these efforts, the AI Safety Institute Consortium was established by the National Institute of Standards and Technology in February. This collaboration, involving over 200 organizations, aims to establish safety guidelines for AI systems, particularly in the areas of generative AI and general AI development and adoption.
According to experts, the United States lags behind several other countries, including the European Union, in terms of AI regulation. In March, the EU implemented a comprehensive new law that governs AI across its 27 member states, placing pressure on the U.S. to catch up.
The EU law includes safeguards for general-purpose AI, limitations on law enforcement’s use of biometric identification systems, prohibitions on online social scoring, AI manipulation, and exploitation of user vulnerabilities. Additionally, it grants consumers the right to file complaints and receive meaningful explanations from AI providers.
In a related article, David Brin, author of the Uplift series, explores ways to prevent an artificial intelligence apocalypse and highlights the importance of responsible AI development.