In a significant move towards enhancing the regulatory landscape for artificial intelligence (AI) in the United States, over 60 commercial organisations, academic institutions, and non-profit groups have collectively urged Congress to authorise the establishment of a US AI Safety Institute within the National Institutes of Standards and Technology (NIST). This collective appeal aligns with bills already introduced in both chambers of Congress, Senate Bill 4178, the Future of AI Innovation Act, and House Bill 9497, the AI Advancement and Reliability Act. These legislative proposals aim to formalise the creation of a NIST-led AI centre focused on advancing AI through research, standards development, and fostering partnerships between public and private sectors.

The Senate bill, which is supported by a bipartisan group of senators including Maria Cantwell (D-Wash.), Todd Young (R-Ind.), and John Hickenlooper (D-Colo.), seeks to officially establish the institute that is already operating within NIST. Similarly, the House bill, sponsored by representatives Jay Obernolte (R-CA-23), Ted Lieu (D-CA-36), and Frank Lucas (R-OK-3), refers to the institute as the Center for AI Advancement and Reliability.

Should the bills receive approval from both legislative chambers, they would be reconciled into a unified piece of legislation to be presented to President Joe Biden for his signature. However, the passage of these bills faces considerable uncertainty amid the current climate of Congressional inaction, with the 118th Congress enacting a significantly low number of laws compared to previous sessions over the past 50 years, as per records from GovTrack.us.

Leading the charge for the legislative push is the coalition spearheaded by Americans for Responsible Innovation (ARI) and the Information Technology Industry Council (ITI). In an open letter released on Tuesday, the coalition emphasized the necessity of supporting NIST's efforts to ensure AI safety to maintain national security and competitiveness. ITI's president and CEO, Jason Oxman, highlighted the importance of authorising the AI Safety Institute to keep the US competitive in the global AI race.

The open letter, signed by notable entities such as Amazon, Google, Microsoft, Lockheed Martin, Palantir, and academic institutions like Carnegie Mellon University, underscores the coalition's focus on national policy objectives and frameworks for evaluating AI systems rather than imposing stringent regulatory limits.

This approach contrasts with California's recent legislative attempt, SB 1047, which sought enforceable measures for AI safety but was vetoed by Governor Gavin Newsom last month due to concerns over its impact on the state economy. The Senate and House bills, however, propose developing voluntary best practices, steering away from the enforceable obligations intended by the California bill.

California Senator Scott Wiener, who authored SB 1047, expressed disappointment over the veto, pointing out the inadequacy of voluntary commitments from industry players in achieving public safety and the lack of binding federal restrictions.

With just over two months left in 2024, whether lawmakers will manage to navigate through the legislative impasse and pass federal AI legislation endorsed by major tech companies remains uncertain.

Source: Noah Wire Services