Congress Should Not Create AI-Specific Regulation, Say Techies

WASHINGTON, July 28, 2023 – Artificial Intelligence experts said that Congress should not make AI-specific legislation to protect against potential harms at a Congressional Internet Caucus Academy panel Friday.

AI harms and risks are already addressed by existing laws, said Joshua Landau, senior counsel of innovation policy at nonprofit advocacy organization the Computer and Communications Industry Association.

Landau urged Congress to write laws that address harms rather than creating laws that specifically regulate AI usage. He warned that differentiating between AI and human crimes will only create loopholes in law that will serve to incentivize unlawful behavior, which in turn will affect where research and development in the industry will go. The exception is laws that delineate liability for harmful actions when AI is involved, he said.

His comments follow an opinion expressed by former Chairman of the Federal Communications Commission, Richard Wiley, on Tuesday who said that now is not the right time to regulate AI and urged lawmakers to slow down in efforts to regulate the technology.

The desire for perfect policy has held Congress back from developing AI regulation, added Evi Fuelle, global policy director at Credo AI. She urged for Congress to implement transparency mandates for both large and small AI companies.

Voluntary commitments will fail to show results if Congress does not mandate them, said Fuelle, referring to the seven AI companies that committed to the White House’s AI goals last week. The commitments included steps to ensure safety, transparency and trustworthiness of the technology.

Nick Garcia, policy counsel at Public Knowledge, cautioned against policies that will call for a pause or halt in AI research and development, saying that it is not a sustainable solution. He also urged Congress to address AI issues without neglecting equally important concerns surrounding social media regulation.

In October, the Biden Administration announced a blueprint for a first-ever AI Bill of Rights that identifies five principles that should guide the design, use and deployment of AI systems in order to protect American citizens. According to the White House, federal agencies have “ramped up their efforts” to protect American citizens from risks posed by AI technology.

In May, Biden signed an executive order directing federal agencies to root out bias in the design of AI technology and protect the public from algorithmic discrimination. Thursday, a House Committee passed legislation that would direct the National Telecommunications and Information Administration to conduct research on accountability measures for AI.