Global: US Senate’s Return Could Accelerate AI Regulation

As summer comes to an end, the march of the generative artificial intelligence (AI) revolution continues unabated. Leading firms like OpenAI, Google, and Anthropic updated their cutting-edge large language models (LLMs) in August. However, progress on the regulatory front for AI technology remains sluggish outside of China.

Change might be on the horizon, though. Senate Majority Leader Chuck Schumer (D-NY) is set to kick off a series of bipartisan “AI Insight Forums” next Wednesday (September 13, 2023). This initiative is expected to pave the way for a more robust framework of AI regulation when the U.S. Senate reconvenes after its summer recess.

Schumer emphasized the importance of addressing AI with the same level of seriousness as national security, job creation, and civil liberties in the 21st century. These AI forums aim to educate American lawmakers about the rapidly advancing potential of AI technology.

The first meeting, scheduled for next week, will include tech sector leaders such as Elon Musk (Tesla and X), Mark Zuckerberg (Meta), Sundar Pichai (Google), Jensen Huang (NVIDIA), Sam Altman (OpenAI), Bill Gates (Microsoft Co-founder), and Satya Nadella (Microsoft CEO), among others with expertise in both technology and policy.

In parallel, Britain has outlined five objectives for its upcoming AI Safety Summit, scheduled for November 1-2. These objectives include creating a shared understanding of the risks associated with frontier AI, fostering international collaboration on AI safety, developing measures to increase AI safety, identifying areas for potential collaboration on AI safety research, and showcasing how safe AI development can benefit global use.

Historically, U.S. lawmakers have taken their time when responding to the emergence of revolutionary technologies with policy frameworks. While the tech industry may prefer this approach, experts, academics, lawmakers, and some tech executives are urging the government to draft regulatory policies swiftly due to the rapid growth of sophisticated generative AI systems.

However, regulating AI poses significant challenges, given its diverse use cases and the need for deep knowledge. The regulation process could take inspiration from the past, where guardrails for new technologies were often based on existing regulations for older technologies that shared similarities.

For example, television regulations were influenced by those in place for radio and telephones, and federal regulation for railroads took about 60 years to develop. Cary Coglianese, founding director of the Penn Program on Regulation, emphasizes that existing regulators may be better suited to oversee certain AI applications.

While it remains uncertain whether the U.S. will pass comprehensive AI regulation, it’s more likely that effective legal frameworks will emerge from individual states’ initiatives. Over 25 states have introduced AI-related legislation this year, with measures passing in 14 of them. This decentralized approach allows states to tailor regulations to their specific needs and provides insights for crafting cohesive national regulations that balance safety, innovation, and economic growth.