Biden-Harris Administration Wants to Keep AI Models Open

WASHINGTON, July 31, 2024 – The National Telecommunications and Information Administration on Tuesday released a “Dual-Use Foundation Models with Widely Available Model Weights” report, which recommended that large artificial intelligence models should continue to to be accessible to everyone.

Previously, the NTIA of the Commerce Department had called for establishing a principle that federal agencies should require independent audits of all AI model classes “that present a high risk of harming rights or safety.” 

The prior report focused on “high risk” AI and called for the use of independent audits to balance public safety and rights with the regulation of AI's development. 

Some of the benefits of these large models, dubbed open foundation models, proponents believe, is that they foster innovation, enhance competition, advance research, and improve accessibility in AI. 

There is insufficient evidence to support restricting the availability of these open models, the NTIA said in the Tuesday report. Still, with the rapid evolution and uncertainty of the AI field, the NTIA recommended that the government continue to monitor and assess risks associated with the technology. 

The NTIA also noted that it had met all of its deadline’s in the Biden Administration’s August 2023 AI executive order. 

Nick Garcia, senior policy counsel at Public Knowledge, praised the recent report and the agency’s recommendation to keep an open environment for AI, calling it a significant step. 

Flexibility is vital for regulating “rapidly changing technologies. A flexible regulatory approach acknowledges the limitations of our current understanding, anticipates the need for adjustments as technology evolves, and positions the government to adapt to new risks that may emerge from AI developments,” said Garcia. 

Following the August 2023 Executive Order on AI, in October the Biden-Harris Administration introduced the AI Bill of Rights with five guiding principles to protect citizens.

The GOP has tried to make political gains against the administration’s AI executive order, claiming it stifles innovation. 

And Silicon Valley insiders have praised the GOP's tech-friendly stance as beneficial for innovation, arguing that Biden’s approach imposed excessive burdens.

On Wednesday, a Senate committee was scheduled to vote on a collection of artificial-intelligence-related bills addressing AI security, private sector standards, education, and privacy.